The job a new accountant walks into this year barely resembles the one I started in. Same title, mostly the same work, but the day itself is different. The work that used to fill a staff accountant’s first two years is the work a tool now does before lunch.

That one fact explains most of how things look for a new grad right now, both the good and the scary.

They come from the same thing, so I’ll go through them together.

What’s working in their favor

The soul-crushing part of year one is shrinking. The endless ticking, the data entry, the basic recs that used to define the first stretch of a career. A new grad spends less of their early years in the trenches and more of it near the work that’s worth doing.

The output ceiling moved up too. A grad can draft a memo, a variance writeup, a first-pass model at a level that used to take a few years to reach. Looking competent early is easier than it used to be.

The stuff that used to sit behind a senior’s time is open now. How do I treat this, where does it go, what does a good version look like. They’re not stuck waiting for someone to have a free thirty minutes. The question gets answered in the moment they have it.

No unlearning either, which I got into a bit last time. For them the tool was part of the work from the very first task, never something they had to bolt on later.

What I’d flag

The catch is that it’s the same fact seen from the other side. The grunt work that’s disappearing was also the gym.

You built your gut feeling doing the boring reps. A thousand recs and tie-outs taught you what a normal one feels like, so the weird one jumps out at you later. If the tool does the reps, that gut doesn’t build on its own. You have to go build it on purpose, which almost nobody does without a reason to.

The trap that comes out of that is fast-but-shallow. The output looks like a senior wrote it, but underneath, the understanding is still year-one, and the gap is invisible, to the grad and sometimes to the manager, right up until something breaks and they’re the one person in the room who can’t tell that it broke.

A model will also hand you a confident wrong answer with a straight face, and a new grad is the worst-equipped person to catch it. A senior would catch it, because they’ve seen the real number enough times that a wrong one feels off. A grad doesn’t have that yet.

The bar moved too. “It took me all day” used to be a fine answer. It isn’t now, because everyone knows what the tool does in twenty minutes. Speed just gets assumed, so cranking out volume doesn’t set a grad apart the way it once did. What does is owning the work, and being able to stand behind it when someone pushes.

I’d keep half an eye on the door itself, too. If the bottom-rung work is what a tool does now, firms start rethinking how many bottom-rung seats they need. I don’t know how that shakes out yet. Worth keeping an eye on.

So what does a new grad do with this?

The advantage is real, but it runs out if you sit on it. The grad who wins spends the time AI gives back building the judgment it can’t.

Pretty concretely: do some things the slow way on purpose, early, while the stakes are low. Work out why a number is what it is, even when the tool already gave you the number. Get one senior to show you the checks they run, the ones that never show up in the final answer.

None of this is exciting. It’s the deliberate version of the reps you’re no longer forced into, and it’s the part that keeps fast from sliding into shallow.

Your move this week

If you’re early in your career: pick one task you’ve only ever done with a tool and do it once by hand, start to finish, slow. It tells you what you understand on your own versus what the tool’s been quietly carrying for you.

If you manage these folks: take your fastest staff and walk them through one check they don’t know to run yet. Show them how you’d catch the error, the part that never shows up in the final number. Twenty minutes, and it’s probably the best mentoring you’ll do all month.

What part of your first couple years taught you the most, and would a new grad even get that experience today? Hit reply and tell me. I don’t really know the answer to that one yet, and I’m trying to work it out.

There is a difference.

A tool is something you pick up when you need it. A system is something you design once and let run. The Accounting Engineer doesn’t just use a stack — they design one. They think about how data moves, where it lives, what connects to what, and what comes out the other side. They build with the end in mind.

That’s what this article is about.

Not a list of software. Not a vendor comparison. A framework for thinking about how to build an AI-native accounting stack that stays lean as the business grows, automates without sacrificing accuracy, and produces the outputs your team needs every month without rebuilding from scratch.

Here’s how it works.

Start with the end in mind

Before you touch a single tool, answer three questions.

1/ What do you need to produce every month? Close commentary. Flux analysis. Financial statements. Reconciliations. Cash analysis. The outputs define the system. If you don’t know what you’re building toward, you’ll end up with a collection of tools that don’t talk to each other.

2/ How lean do you need to stay? A two-person finance team at a Series A company has different requirements than a ten-person team at a Series C. The stack should scale alongside the business without requiring you to add headcount every time the volume goes up.

3/ What needs to be automated and what needs a human? Not everything should be automated. The goal isn’t to remove humans from the process — it’s to remove humans from the work that doesn’t require them. The Accounting Engineer designs that line intentionally. Everything below it gets automated. Everything above it gets their attention.

Answer those three questions first. The rest follows.

The five layers

An AI-native accounting stack has five layers. Each one builds on the one below it. Get the foundation right and everything above it gets better.

Layer 1 — Data Sources

This is where your data lives. ERP, spend management, payroll, billing, and anything else that produces financial data. The key question at this layer isn’t which tool — it’s whether each tool can get your data out cleanly and consistently. Tools that silo data are a liability. Tools that expose it are an asset.

Modern ERPs are increasingly building MCP connections — meaning Claude can reach in and pull data directly without a manual export. NetSuite, Rillet, and Campfire all have MCP. Spend management platforms like Ramp and Brex do too. If your ERP doesn’t expose your data cleanly, that’s a constraint worth solving before you build anything on top of it.

Layer 2 — Semantic Layer

Raw data from multiple systems is messy. The same concept has different names in different tools. Revenue in your ERP might be labeled differently than revenue in your billing system. The semantic layer is where you standardize — define your metrics, create consistent naming, and make sure that when you ask for “revenue,” every system means the same thing.

This layer is often skipped by smaller teams. That’s a mistake. The semantic layer is what makes the data warehouse trustworthy. Without it, you’re building on an unstable foundation.

Layer 3 — Data Warehouse / Data Lake

This is your single source of truth. All of your financial data — from every source, through the semantic layer — lives here, clean and ready. When Claude pulls data for a flux analysis, it’s pulling from here. When you build a dashboard, it’s reading from here. When an auditor asks for the methodology, you can point here.

The warehouse is what turns a collection of disconnected tools into a system. It’s the layer that makes everything else possible.

Layer 4 — Claude

This is where the work happens. Claude connects to your warehouse via MCP and pulls live data directly — no exports, no copy paste, no manual steps in between. Chat to design what you want to see. Cowork to execute complex multi-step workflows. Code to build exactly what you need. Projects to save the context so you’re not starting from scratch every month.

The quality of what Claude produces is directly tied to the quality of the layers underneath it. Clean data, well-defined metrics, a reliable warehouse — Claude takes that and produces outputs that are structured, consistent, and ready to use. The better the foundation, the better the output.

Layer 5 — Monthly Deliverables

This is what your stack produces. Every month, consistently, without rebuilding from scratch.

Flux analysis — month over month variance with commentary, pulled directly from your warehouse, generated by Claude in the format your CFO expects.

Cash analysis — inflows, outflows, variance to forecast. Claude connects, pulls the data, produces the output.

Financial statements — P&L, balance sheet, cash flow. Structured, reviewed, ready.

Reconciliations — account, bank, intercompany. The agent runs the process. You review the output.

Close commentary — CFO-ready narrative on the period, written in the tone and format you’ve defined in your Project.

These aren’t aspirational. They’re workflows accountants are running right now. The stack makes them repeatable.

What this looks like at a lean company

Here’s what this stack looks like in practice for a lean finance team at an early-stage company.

The ERP is Rillet/Campfire — AI-native, MCP-connected, built for teams that need to move fast without sacrificing accuracy. Spend management is Ramp/Brex — MCP connection means Claude can pull transaction data directly. Payroll and billing feed in. The semantic layer is lightweight but defined — consistent account naming, consistent metric definitions. The warehouse is the layer that ties it all together.

Every month, Claude connects via MCP, pulls the current period data, compares to prior period, and produces flux analysis, cash analysis, and close commentary — all in a format that’s been defined once and reused every month. The close that used to take days now takes a fraction of that. Not because the work disappeared. Because the system handles the work that doesn’t require a human.

The accountant who built that system isn’t spending time on exports and formatting. They’re spending time on the analysis that actually moves the business.

That’s the Accounting Engineer.

Three things to avoid

The biggest mistakes when building an AI-native stack aren’t technical. They’re structural.

1/ Buying tools before defining the architecture. The natural instinct is to start with the tool — find the best ERP, find the best spend management platform, find the best AI. But tools without architecture are just more things to maintain. Start with the end in mind. Define what you need to produce. Then build toward it.

2/ Automating for automation’s sake. Not every process should be automated. The ones that should are the ones that are repetitive, rule-based, and high-volume. The ones that shouldn’t are the ones that require judgment, context, and human oversight. Build the line between them intentionally.

3/ Building something only you can run. The best accounting systems aren’t dependent on one person. They’re documented, organized, and transferable. If the only person who can explain your AI workflows is you — that’s a risk, not an asset. Build like someone else will need to run it.

Building for the future

The accounting profession is in a building phase. The teams that come out of this period ahead aren’t the ones with the most tools — they’re the ones that built the most intentional systems.

Stay lean. Build with the end in mind. Design the stack so the data flows, the outputs are consistent, and the work that doesn’t require a human doesn’t get one.

That’s the Accounting Engineer’s job now.

Go build it.

Here’s my question to you:

What’s the biggest gap in your current stack — the layer that’s holding everything else back? Drop it in the comments. I read every one.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

Every other week someone sends me an article about which jobs AI is going to wipe out, and accounting is almost always on the list. Data-heavy, rules-based, repetitive. The kind of work a model can supposedly do in its sleep.

I think they’ve got it backwards.

The skills people point to as the reason accountants are at risk are the same skills that make us the most prepared profession for this shift. More prepared than the marketers, the engineers, or the consultants. We just don’t talk about it that way because we’re too busy closing the books.

So let’s go through it.

You already know systems

AI doesn’t do much on its own. It has to plug into something: a GL, an ERP, a reporting layer, a pile of CSVs that somebody exports every month. The value shows up when the tool connects to the actual plumbing of a business.

Accountants live in that plumbing. You know where the data sits, how it moves, where it breaks, and who touches it before it gets to you. You know that a certain subledger doesn’t tie to the GL until the third business day, and why. That kind of knowledge isn’t written down anywhere a model can read it. It lives in your head.

Most people trying to build with AI right now are stuck because they don’t understand the systems they’re trying to automate. You don’t have that problem. You’ve been mapping how data flows through a company for your whole career.

You already live in data

A model will hand you a number and say it with total confidence. Sometimes the number is right. Sometimes it’s made up. The model has no idea which.

You do. Or at least you know how to find out.

Professional skepticism is drilled into accountants from day one. Tie it out. Trace it back. Does this even make sense? We don’t trust an output until it agrees with something else. That instinct, the refusal to take a number at face value, is the single most valuable habit you can bring to working with AI, because these tools hallucinate, and the people who catch it are the ones trained to never trust the first answer.

Interpreting data is the other half. A model can produce a variance table in seconds. Knowing which variance matters most, what’s material, what the CFO is going to ask about: that’s judgment only you have context on.

You already talk to everyone

This one gets overlooked, and I think it might be the most important.

Accountants sit in the middle of the business. In a given week you might talk to the CEO, an auditor, a banker, the sales team, a vendor, and the tax authority. You translate between the people who make the numbers and the people who read them. You’re the connective tissue.

The bottleneck in AI adoption usually isn’t the technology. It’s finding someone who understands the business problem, understands the data, and can explain the whole thing to a tool clearly enough to get something useful back. It turns out accountants are weirdly good at it, because we already spend our days translating messy reality into structured output for an audience.

You already think in process

Stop and look at what a reconciliation really is. Or a month-end close. Or a recurring journal entry.

They’re workflows. Step one, step two, check this against that, flag the exception, document it, move on. You’ve been writing algorithms this whole time, you just called them procedures and stored them in a checklist instead of code.

That’s the same shape as building an automation. The accountants I’ve watched pick up AI fastest aren’t the most technical ones. They’re the ones who already think in repeatable steps, because that’s how they were trained to close the books without missing anything.

The rules thing helps too. We work inside GAAP, inside the tax code, inside our own internal controls. Constraints everywhere. Models happen to work best when you can define the rules and the output you want, and defining rules is most of the job already.

So where does that leave you

I’m not going to pretend the next few years won’t be uncomfortable. Some tasks are going to get automated, probably faster than we’d like. If your whole role is keying in invoices, that part really is at risk, and I’d be lying if I said otherwise.

But the task was never the whole job. The judgment, the systems knowledge, the skepticism, the ability to sit between the data and the people who need to read it, all of that stays. If anything, AI makes it more valuable, because now you can do the boring 80% in a fraction of the time and spend the rest where you add something real.

The accountants who win the next decade won’t be the ones who memorize how a transformer works under the hood. They’ll be the ones who looked at a slow, painful workflow, opened a tool, and tried to build something better. Builders, not bookkeepers.

You’re better positioned for this than almost anyone. Might as well use it.

Your move this week

Pick one task you do every month that follows the same steps every time. A recon, a recurring entry, a report build, whatever you could do half asleep.

Write out the steps the way you’d explain them to a new hire. Then paste that into Claude or ChatGPT and ask it to walk through the process with you using this month’s real data.

You’re not automating anything yet. You’re just seeing how close the tool gets when you hand it the process you already know cold. That’s the whole first step, and it takes about twenty minutes.

What’s the first workflow you’d hand off to AI if you trusted it to get it right? Hit reply and tell me, I read every response.

Chat to converse and design what you want to see. Code to build exactly what you want. Cowork to be your multi-task assistant. (Vol. 4)

The question is — where is the data coming from?

More often than not you’re manually downloading it. Exporting a CSV. Copying a table. Pasting it in. Every single time.

You’re leaving time saved on the table. You’re leaving automation on the table.

That’s where MCP comes in.

What is MCP?

MCP stands for Model Context Protocol. Forget the name — here’s what it actually means for an accountant.

It’s a bridge between Claude and your actual systems. Instead of exporting data, copying a table, or manually typing numbers into a prompt — Claude connects directly to your tools and pulls the data itself. Live data. No copy paste. No export. No manual step in between.

Think about what that means in practice. The flux analysis from Vol. 5 — pulling chart of accounts, pulling period balances, comparing periods, generating commentary — none of that required a single manual data export. MCP pulled it directly from the system.

That’s the shift. Chat, Code, and Cowork design what you want Claude to do. MCP gives it the data to do it with.

Where does MCP connect?

This is where it gets practical for accountants. There are three categories worth knowing.

Your ERP

Most modern ERPs have built MCP connections. NetSuite, Rillet, and Campfire all have MCP — which means instead of exporting a report, Claude can reach directly into your chart of accounts, pull period balances, and work with live data. No file. No download. No manual step.

If your ERP has an MCP connection, this is where you start.

Spend Management

Same story on the spend side. Ramp and Brex both have native MCP connections. Claude can pull transaction data, card activity, and spend by category directly — without you touching a CSV. Ask Claude to analyze last month’s spend by vendor, flag anything over a threshold, or compare to prior period. It goes and gets the data itself.

Admin and workflow tools

Beyond the financial systems there’s a growing library of MCP connectors for the tools accountants use every day.

• Slack — Claude can pull messages, threads, and context from conversations your team has already had.

• Email — surface relevant threads without switching windows.

• Google Drive — pull documents and spreadsheets directly into the conversation.

Calendar tools, project management systems, and more are being added constantly. The pattern across all of them is the same. Connect once. Claude pulls what it needs when it needs it.

What this looks like in practice

Here’s the clearest way to understand the difference MCP makes.

Without MCP: Open ERP. Export data. Save file. Upload to Claude. Describe the format. Ask Claude to run the analysis.

With MCP: Tell Claude what you need. It connects, pulls the data, runs the analysis.

Same output. Fraction of the friction.

How to get started

Getting connected is simpler than it sounds.

Step 1 — Open Claude desktop and go to Settings. Find the Connectors section.

Step 2 — Browse the available connectors. Look for the tools you already use — your ERP, your spend management platform, your admin tools.

Step 3 — Connect and authorize. Claude will ask for the permissions it needs. It only sees what your account is allowed to see — the same permissions you already have.

Step 4 — Test it. Open a conversation and ask Claude something that requires data from that system. See what it pulls back.

Step 5 — Build from there. Once you’ve seen it work once, you’ll immediately start thinking about every other workflow where you’re manually moving data that MCP could handle automatically.

The bigger picture

MCP is what turns Claude from a chat tool into an accounting system.

Everything from Vol. 1 through Vol. 5 gets more powerful with MCP underneath it. Better prompts on live data. Projects that pull real numbers. Cowork workflows that run without manual steps in between.

The accountants connecting their tools right now are operating at a completely different level than the ones still copying and pasting. Not because they’re more technical. Because they took the time to connect once and let the system do the rest.

Vol. 7 is next. Keep building.

Here’s my question to you:

What system would you most want Claude to connect to directly? Drop it in the comments — I read every one.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

They build a great workflow. Get a great output. Close the tab. Next month they open a brand new chat and start from scratch — same context, same instructions, same examples, rebuilt from zero.

Every. Single. Month.

This issue is about fixing that. There are two ways to do it. Both work. Pick the one that fits how you work.

The problem with starting fresh

Every time you open a new chat, Claude doesn’t know anything about you. It doesn’t know your chart of accounts. It doesn’t know your format. It doesn’t know your CFO’s communication style or what threshold you use for material variances.

So you paste it all in. Again. Or you skip it and get generic output. Again.

The goal is to build the context once and reuse it every month. Here are the two ways to do that.

Option 1 — The saved Cowork chat

This is the simpler path and it works.

Cowork is Claude working through a multi-step task autonomously within a single conversation (Vol. 4). The key insight most people miss — you don’t have to start a new chat every month. You can go back into the same conversation and Claude already has the full history. The context is there. The workflow is there. You pick up where you left off.

Here’s what that looks like in practice with a real workflow.

Every month I run a flux analysis. Using Claude’s MCP connection I pull chart of account names directly from our system. Then by period I pull the balances for each account. Claude compares current period to prior period — calculating dollar and percentage change for every line. Then it pulls the data together and generates commentary explaining the material movements.

One workflow. Chart of accounts pulled. Balances compared. Variance calculated. Commentary written.

The first time I set this up I built the context — my role, my company, my format, my materiality threshold. Now every month I go back into that same chat. Claude already knows all of it. I connect via MCP, pull the current data, and the workflow runs.

Best for: Individual workflows you own yourself. Recurring monthly tasks where you’re the only one using it. Getting started quickly without a lot of setup.

The one limitation: If you accidentally close the chat, start a new one, or want someone else on your team to run the same workflow — you’re rebuilding from scratch. The context lives in the conversation history, not above it.

Option 2 — Claude Projects

Projects is the more structured path. Same goal — build once, use every month — but the context lives at the workspace level, not inside a conversation.

A Project is a persistent workspace where you write standing instructions and upload reference documents once. Every conversation you open inside that Project — whether it’s your first or your fiftieth, whether it’s you or someone on your team — starts with all of that already loaded. Claude already knows everything before you type a single word.

Three things go inside a Project:

Instructions — who Claude is, what context it has, how it should behave (Vol. 3). Your persistent CO-STAR framework baked in permanently (Vol. 2). For a flux analysis Project this might include your role, your company’s industry, your materiality threshold, your CFO’s preferred tone, and how you want variances presented.

Knowledge — documents you upload once and reference forever. Your chart of accounts. A prior month’s flux as a format template. Any standing context that would otherwise take five minutes to paste in every conversation.

Conversations — every chat inside the Project starts from the same foundation. New conversation, same context. The history doesn’t have to carry it — the Project does.

Best for: Workflows you want to share with a team. Workflows that rely on uploaded reference documents. Building a clean organized library of named workspaces instead of hunting through chat history. Scaling beyond yourself.

Which one should you use?

Honestly — it depends.

If you’re a solo accountant running your own monthly workflows and you’re comfortable going back into the same chat every month — the saved Cowork approach works. It’s simpler to set up and gets you the same result.

If you want something more organized, more shareable, more reliable regardless of what happens to any individual conversation — build a Project. The setup takes longer upfront but the infrastructure is cleaner and scales better.

Both beat starting from scratch every month. That’s the point.

How to build your first Project

Step 1 — Open Claude desktop and find the Projects section in the left sidebar.

Step 2 — Create a new Project and name it for the job it does. “Monthly Flux Analysis.” “Close Commentary.” “Cash Analysis.”

Step 3 — Write your instructions. Start simple: who Claude is, what company context it needs, what standard it holds, how you want the output formatted. Refine over time.

Step 4 — Upload your knowledge. At minimum: your chart of accounts and one prior output as a format template.

Step 5 — Open your first conversation inside the Project. Test it. Refine the instructions based on what’s missing. It gets better every time you use it.

The compounding effect

Whether you go with a saved Cowork chat or a Project — the principle is the same.

Build the context once. Every month after that you’re not starting from scratch. You’re building on something that already works.

The first month takes the most effort. The second month is faster. The third month faster still. By month six you have a workflow that produces consistent structured output faster than anything you’ve ever done manually.

That’s the compounding effect of building once instead of rebuilding every month.

Pick your most repetitive monthly workflow. The one where you spend the first ten minutes re-explaining context that hasn’t changed. Set it up — as a saved Cowork chat or a Project. Run it once. See what the second month feels like.

Vol. 6 is next. Keep building.

Here’s my question to you:

What’s the first workflow you’re going to stop rebuilding every month? Drop it in the comments — I read every one.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

The accounting profession is in a building phase right now. Workflows are being automated. Closes are getting faster. Commentary that used to take hours is being produced in minutes. The tools are here, the results are immediate, and the momentum is real.

But here’s the question almost nobody is asking while they’re building:

Can you defend it?

Not “does it look right” or “does it pass the gut check”. But, can you sit in front of an auditor, a CFO, or a board member and walk them through exactly what happened, what the AI did, what you reviewed, and why you’re confident in the output?

That’s the auditability question. And the accountants building with AI right now need to be thinking about it from day one – not after the audit work has begun.

What is an AI agent?

An AI agent isn’t just a chatbot that answers a question. It’s a system that executes multiple steps, makes decisions along the way, and produces an output — often without accountants watching and being involved in the output step by step like they have in the past.

Said another way, when you ask Claude a question in Chat, you’re conversing. You see every response. You adjust and review the output. Then you make the decision of what you do with it.

An agent is different. You assist with the initial setup and tweaks but once you go live, it’s hands off. You give the agent a task (e.g. run analysis, process data, generate report) and it works through the steps autonomously. The output arrives and the steps that produced it aren’t always visible — and that is exactly where the auditability question lives.

The more powerful and reliable the agents, the higher the audit risk.

And that risk doesn’t manage itself.

Why this matters

For decades the profession has built audit standards around human work. A human has prepared reconciliations – there’s a paper trail, a reviewer, a sign-off. A human has written the variance commentary – all with preparer, reviewer and other records of the process.

AI agents are being treated as task replacements. The agent did the reconciliation. The agent wrote the commentary. Task complete. But that framing misses something critical.

In terms of auditability, these workflows deserve the same level of scrutiny we’d apply to any automated process in a financial system. Except there’s a human element that makes it even more important — because the output is only as good as the person who built it. The instructions they gave. The context they provided. The review they built in. The judgment calls they made along the way.

That’s what makes the Accounting Engineer’s role so important here. You built it. That means you own it. And owning it means being able to defend every step of it when the auditor arrives.

That’s the standard. Everything else in this article is about how to meet it.

Why assertions still matter

If you’ve spent any time in public accounting or working closely with auditors, you know the assertions. Existence. Accuracy. Completeness. Cutoff. Presentation. They’re the foundation of how auditors gain assurance over financial statements — the questions they’re asking when they pull a sample, test a control, or trace a transaction.

They didn’t go away when AI arrived. If anything, they became more important.

Here’s why: when a human prepares a reconciliation, an auditor can sit down with that person and ask them to walk through every step. Where did the data come from? How did you identify the reconciling items? What did you do when something didn’t tie? The human can answer. The evidence exists. The judgment calls can be explained.

When an AI agent runs that same reconciliation — who answers those questions?

That’s not a reason to stop building. It’s a reason to build differently.

The assertions as a lens

You don’t need to go deep on every assertion to make this point — but it helps to see them as a lens for evaluating any AI workflow you’re building or deploying.

Existence — did this transaction actually occur, and can you prove the agent processed it correctly? If an agent is touching transaction data, the auditor is going to want to know the process was reliable and the output reflects reality. What’s your evidence?

Accuracy — is the amount right? AI can produce outputs that look correct and aren’t. Without a defined verification step built into the workflow, a plausible-looking wrong number can move through a process unchallenged. Accuracy doesn’t assume itself just because AI produced the output.

Completeness — is everything that should be there, there? An agent following instructions doesn’t know what it doesn’t know. If something falls outside the parameters you set, it may not flag it. Completeness requires the person who built the workflow to have thought through the edges — what could be missing and how would we know?

Cutoff — is it in the right period? Period-end judgment calls are some of the most nuanced decisions in accounting. An agent doesn’t have the context for those calls unless you’ve built it in explicitly. Cutoff errors are one of the most common audit findings — and they don’t get less likely just because AI is involved.

Presentation — is it presented correctly? An agent presents based on the instructions and context it was given. If the context is incomplete or the instructions are ambiguous, classification errors follow. The auditor is going to ask how you know it’s right.

The question underneath all of them

Every assertion comes back to the same underlying question: can you demonstrate that the process produced a reliable output — and can you show your work?

For a human process, showing your work is second nature. For an AI workflow, it has to be intentional. It has to be built in. Because if you can’t answer the assertion questions, the auditor can’t gain assurance — and without assurance, the work doesn’t stand.

This is what is going to separate the successful audits from the ones that result in deficiencies.

What Goes Wrong Without Auditability

Let’s get specific. Because the risk here isn’t theoretical — it’s the kind of thing that shows up in audit findings, management letters, and uncomfortable conversations with your CFO.

Here are three scenarios where the absence of an audit trail becomes a real problem.

1/ The auditor asks why — and you can’t answer.

You built a workflow that generates variance commentary. The numbers look right. The narrative reads well. The CFO signed off. Then audit fieldwork starts and the auditor pulls the workpaper. They want to understand the methodology. How was the variance calculated? What data did it pull from? How do you know it captured everything?

You ran a prompt. You reviewed the output. It looked right so you moved forward.

That’s not a methodology. That’s a task. And there’s a significant difference between the two when an auditor is sitting across the table asking questions you can’t answer.

2/ The output was wrong and nobody caught it.

AI produces plausible-sounding outputs. That’s one of the things that makes it so useful — and one of the things that makes it genuinely dangerous in a financial context if there’s no structured review layer.

Hallucination is real. Gaps in context produce gaps in output. An agent that wasn’t given the right parameters will work within the ones it was given — and produce something that looks complete when it isn’t. Without a defined verification step built into the workflow, a wrong number moves through unchallenged. It makes it into the report. Into the board deck. Into the filing.

By the time someone catches it, the damage is done.

3/ Nobody owns it.

This is the one that keeps coming up and it doesn’t have a clean answer yet.

When an AI agent makes a decision in a financial workflow — who is responsible for that decision? The accountant who built the workflow? The one who reviewed the output? The one who signed off on the report?

Without clear ownership built in from the start, the answer defaults to nobody. And nobody is not an acceptable answer when a deficiency gets written up, when a restatement conversation starts, or when someone is trying to understand how an error made it through the process.

The agent didn’t make the decision. The person who deployed the agent made the decision. That accountability has to be explicit — in the workflow, in the documentation, and in the mindset of whoever built it.

These aren’t edge cases. They’re the natural consequence of moving fast without building the right infrastructure around what you’re creating. The next section is about what that infrastructure actually looks like.

Building It Right From Day One

The good news is that auditability isn’t complicated. It’s not a separate workstream or a box you check at the end. It’s a mindset you bring to the build — and if you bring it from the start, it becomes part of the workflow rather than something you retrofit after the fact.

Here’s what building it right actually looks like.

1/ Design with the assertions in mind.

Before you deploy any AI workflow into a live financial process, ask yourself the assertion questions. Can I demonstrate existence? Can I prove accuracy? Can I show completeness? Can I defend cutoff? Can I explain presentation?

If you can answer those questions before you build, you’ll build differently. You’ll think about what evidence needs to exist, where the human review points need to be, and what documentation needs to accompany the output. You’re not building a prompt. You’re building a process. And processes have standards.

2/ Build human checkpoints at the right moments.

AI doing the work doesn’t mean humans are out of the picture. It means humans are in the picture at the right moments — not reviewing everything, because that defeats the purpose, but reviewing the things that matter most.

A human checkpoint isn’t “I looked at the output and it seemed fine.” It’s a defined review step tied to a specific assertion. For a flux analysis — did a human confirm the variance drivers are accurate and complete before this goes to the CFO? For a reconciliation — did a human verify that the items flagged by the agent are the right items and nothing was missed?

The checkpoint needs to be documented. Who reviewed it, when, and what they confirmed. That documentation is your audit trail.

3/ Document the methodology.

Every AI workflow that touches a financial process needs a documented methodology. What instructions were given to the agent? What context was provided? What data did it have access to? What was it asked to produce?

This doesn’t have to be a lengthy document. It has to be clear enough that someone unfamiliar with the workflow could pick it up and understand what was done, why it was done, and on what basis. That’s the same standard we’ve always held human work to. It applies here too.

If you built it and you can’t explain it, it’s not ready.

4/ Own the output.

The agent isn’t the accountant. The person who designed the workflow, set the instructions, and deployed it into the process owns the output. Full stop.

That ownership is what makes everything else defensible. The auditor doesn’t need to gain assurance over the AI — they need to gain assurance over the process. And the process has a person behind it. That person needs to be able to stand behind it.

5/ Test it before you trust it.

Before any AI workflow goes live in a financial process, run it the way an auditor would test it. Pull one transaction. Trace it end to end. Follow every step from input to output. Can you explain every decision the agent made along the way? Can you tie the output back to the source data?

If you can do that — if you can walk an auditor through the workflow the same way you’d walk them through a manual process — it’s ready. If you can’t, it isn’t.

This isn’t about slowing down the build. It’s about making sure what you built actually holds up. The goal is an AI workflow you can defend with the same confidence you’d defend any other control in your process.

Getting Auditors on Board

This is the question underneath everything else in this article: how do auditors actually gain assurance over AI agents?

The honest answer is the profession is still working through it. The standards are evolving. The guidance is catching up. But the framework for how auditors gain assurance hasn’t fundamentally changed — and that’s actually the most useful thing to understand right now.

Auditors gain assurance the same way they always have. They understand the control environment. They test the controls. They verify the outputs. They ask questions and expect answers. The fact that an AI agent is involved in the process doesn’t change what they need — it changes what you need to give them.

And right now, most AI workflows aren’t built to give them anything.

Get your own house in order first.

Before you think about the auditor, think about yourself. Can you sit down today and walk someone through every AI workflow you’ve deployed into your financial process? Do you know what instructions each agent was given? Do you have the methodology written down? Do you know where the human review points are and what they confirmed?

If the answer is no — start there.

Organize your agents and your Projects. Document what each one does, what it was built to produce, and what the review process looks like. Build a simple log of the workflows you’ve deployed — what they are, when they were implemented, and who owns them. This doesn’t have to be complicated. It has to exist.

The goal is simple: if someone walked in tomorrow and asked you to explain your AI workflows, you could do it. You could pull the documentation, walk them through the methodology, show them the review steps, and answer the assertion questions with confidence.

That’s the foundation. Everything else builds on top of it.

Then inform your auditors.

Once your house is in order, bring your auditors into the conversation — not to co-build the workflow, not to get their sign-off before you proceed, but to make sure there are no surprises when fieldwork starts.

A brief conversation early goes a long way. Here’s what we’ve been building. Here’s how we’ve documented it. Here’s where the human review lives. Here’s what the output looks like and how we verify it.

That conversation doesn’t need to be extensive. It needs to happen before the audit — not during it. An auditor who has context going into fieldwork is a very different experience from an auditor who is seeing your AI workflows for the first time while they’re trying to gain assurance over your financials.

You’re not asking for permission. You’re giving them the visibility they need to do their job. That’s a professional courtesy that will pay off every time.

The profession is defining this in real time.

There are no perfect answers here yet. The standards will catch up. The guidance will get more specific. Best practices will emerge. But right now, in this building phase, the accountants who are thinking about auditability — who are organizing their workflows, documenting their methodology, building human review into the process, and owning their outputs — are the ones writing the first draft of what responsible AI use in accounting actually looks like.

That matters. Not just for your audit. For the profession.

Build With the End in Mind

The profession is in a building phase. The tools are real, the results are real, and the momentum isn’t slowing down. Accountants are automating processes that used to take days. They’re producing outputs that used to require teams. They’re moving faster than anyone expected.

And most of them aren’t thinking about what happens when the auditor arrives.

That’s the gap. Not the technology. Not the tools. The mindset.

AI agents are being deployed into financial workflows the way a task gets handed off — here’s what I need, go do it. But unlike handing a task to a person, the agent can’t sit in a room and explain its work. It can’t walk an auditor through every step. It can’t answer the questions that fieldwork generates.

That’s your job. You built it. You own it. You have to be able to defend it.

The Accounting Engineer who understands this builds differently from the start. They’re not just asking “what can this agent do” — they’re asking “how do I build this in a way I can stand behind.” They design with the assertions in mind. They document the methodology. They organize their workflows before anyone asks them to. They build the human checkpoints in. They own the output.

That’s not a constraint on building. That’s what building responsibly looks like.

The audit is coming. The standards are evolving. The profession is figuring this out in real time. The accountants who bring auditability into the build from day one — who treat their AI workflows with the same rigor they’d apply to any other control in the process — are the ones who will be ready.

The ones who don’t will find out the hard way.

Build with the end in mind. Always.

Here’s my question to you:

How are you thinking about auditability in your AI workflows right now? Are you building the audit trail in from day one — or figuring it out as you go? Drop it in the comments. This is a conversation the profession needs to be having.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

The fear makes sense. I’m not going to dismiss it.

You went to school for this. You passed the exams. You put in the hours. And now every other headline is telling you the thing you built your career around is about to be automated away. That’s a lot to sit with, especially if you’re early in your career and still figuring out what the next ten years are supposed to look like.

But here’s where I think the fear goes wrong.

Most accountants are afraid of being replaced by AI. What they should actually be thinking about is what happens to the accountants who learn to use it. Because that’s the story nobody is telling loudly enough.

The accounting career path most of us know has always looked the same.

Staff. Senior. Manager. Director. Maybe controller, maybe CFO if the timing works out and you’re willing to put in the years. You know the ladder. You’ve probably already figured out where you are on it.

That path isn’t going away. People are still climbing it.

But something underneath it is shifting. And if you’re not paying attention, you’ll miss the part where a completely different set of doors opens up. Doors that didn’t exist five years ago. Doors that don’t have clean titles yet. Doors that are already open for the accountants willing to walk through them.

Let me be specific, because “AI creates new opportunities” is the kind of sentence that sounds good and means nothing.

Here’s what’s actually happening right now.

The controller who automated her close process spends the last week of the month doing analysis instead of chasing reconciling items. Her CFO started pulling her into board prep. That wasn’t in her job description six months ago. It is now.

The senior accountant who built a cash flow dashboard that updates daily didn’t get a new title. But his CFO opens it every morning and asks him questions. That visibility didn’t come from his credentials. It came from building something.

The Big Four associate two years out who learned to use AI effectively is doing work that used to sit on a manager’s desk. Not because she replaced anyone, but because the work wasn’t getting done at all. Now it is. And people notice.

So what does the new path actually look like?

I think it branches somewhere most people aren’t expecting. And it branches early.

The first branch is the path you already know. You get better at the work, you manage people, you move up. AI makes you faster and more thorough at every step. You use it the way you use Excel. A tool that makes the existing job better. Solid path. Nothing wrong with it.

The second branch is where it gets interesting.

The second branch is the accountant who decides, maybe in year one, maybe in year five, to stop just doing the work and start building the systems around it. The person who automates the close, builds the reporting infrastructure, designs the workflows the entire team runs inside of. The Accounting Engineer.

That person doesn’t move up the traditional ladder. They step off it and into a role that doesn’t have a clean title yet, pays differently, and carries a kind of leverage the old path never offered.

One person with deep accounting knowledge and the ability to use AI to build systems can do what used to take a team. I’ve seen it. The tools are already there.

If you’re a new hire or a recent grad, I want you to hear this part specifically.

You are starting at the best possible moment.

I know it doesn’t feel that way. You’re learning the basics, figuring out the culture, probably being handed the work nobody else wants and told to figure it out. That’s all real.

But you don’t carry the technical debt.

The accountants who’ve been doing this fifteen years built their habits in a world without these tools. Relearning how to work is harder than learning how to work in the first place. You get to build the right habits from the start. You get to learn prompting at the same time you’re learning the close. You get to ask “why does this work this way?” before you’ve accepted that it just always has.

That’s a bigger advantage than you probably realize right now.

And for the manager, the controller, the person ten years in reading this with a little skepticism:

The AI that’s coming for your job isn’t the AI that exists right now.

What exists right now makes good accountants better and fast accountants faster. It doesn’t replace judgment. It doesn’t replace experience. It doesn’t replace the relationship with your CFO that took years to build.

What it does is clear away the repetitive work that was crowding all of that out.

The accountants I’ve seen most threatened by AI are the ones whose value was almost entirely in volume. How much they could process, how fast, how accurately. That work is changing. Probably a lot.

The accountants I’ve seen most energized by AI are the ones who always wanted to do more than process. Who had opinions about the numbers, who wanted to advise instead of just report, who were frustrated that the close took ten days when it should take three.

For that person, this is the best career moment in a generation.

The path doesn’t disappear. It expands.

The staff accountant who starts building becomes something the profession doesn’t have a word for yet. The controller who automates the close becomes the person the CFO can’t operate without. The senior who learns these tools effectively moves faster than the traditional timeline ever allowed. Not by gaming anything, but by delivering work that used to require someone two levels above them.

That’s happening right now. At companies your size. In teams that look a lot like yours.

The career path nobody drew for you is being built by the people who decided to pick up the tools and start.

You can be one of them.

This week’s action item:

Write down the one task in your job that frustrates you most because of how long it takes or how often you repeat it. Just one. Don’t overthink it.

Then open Claude or ChatGPT, describe the task in detail, and ask: “How could I use AI to make this faster or eliminate it entirely?”

Don’t worry about whether the answer is perfect or immediately usable. The goal this week is just to see what’s possible. That’s how it starts for most people. One conversation, one task, one small look at what building could feel like.

Drop your task in the comments. Let’s figure it out together.

But that’s like buying a Swiss Army knife and only ever using the bottle opener.

The same AI you’ve been using for chat can write and run code, build tools, and execute multi-step workflows — all without a computer science degree, a developer on staff, or an enterprise budget.

This is where the Accounting Engineer starts to take shape. Not just someone who uses AI — someone who knows which tool to use and when.

Today we’re covering three modes: Chat, Cowork, and Code. Same AI. Completely different capabilities. By the end of this issue you’ll know what each one is, when to use it, and what it looks like in an accounting workflow.

Let’s go.

Chat

Chat is where everyone starts — and for good reason. You type, it responds. Back and forth, in real time, like a conversation with someone who has read everything.

What it is: the conversational interface. You bring the question, the task, or the problem. Claude brings the knowledge, the drafting ability, and the patience to work through it with you.

When to use it: anything that benefits from a back-and-forth exchange. Drafting variance commentary. Summarizing a contract. Explaining a complex accounting standard in plain language. Preparing talking points for a board meeting. Writing the close narrative your CFO actually wants to read.

Here’s what a Chat prompt looks like in practice:

That’s Chat at its best. Specific context, clear task, defined audience. The output comes back ready to use or one refinement away from being sent.

The thing most accountants don’t realize: Chat isn’t just for one-off tasks. The conversation is the tool. The longer and more specific the thread, the better the output gets. Vol. 3’s framing techniques — establishing a role, providing an example, asking to be asked — all live here. If you’re not using those yet, go back and give it a read.

Chat is the foundation. But it’s not the ceiling.

Cowork

Chat is a conversation. Cowork is something else entirely.

Cowork is Claude working alongside you on a multi-step task — not just responding to one prompt but executing a workflow. You hand it a complex job. It breaks it down, works through the steps, and comes back with something you couldn’t have produced as fast on your own.

This is where AI stops feeling like a tool and starts feeling like a team member.

What it is: Claude operating autonomously across multiple steps within a single task. Think of it like handing a project to a highly capable colleague and saying "here's what I need" — and coming back to a finished product instead of a blank page.

When to use it: complex tasks that would normally require moving between multiple tools and sources. Anything where the output has multiple parts. Anything where you’d normally spend an hour pulling pieces together before you could even start the real work.

Here’s what a Cowork task looks like in practice:

You’re about to renew a vendor contract. Instead of spending two hours reading through the agreement, cross-referencing terms, and drafting a memo from scratch, you hand it to Claude:

One prompt. Three deliverables. What used to take a morning now takes minutes.

The thing most accountants don’t realize: most people never get here because they stop at Chat. Cowork requires you to think differently about how you assign work — not “what do I need to ask” but “what outcome do I need and what would it take to get there.”

Code

This is where most accountants stop reading. Don’t.

You don’t need to know how to code to use Claude’s code capabilities. You need to know the problem well enough to describe it. That’s the accountant’s advantage — you understand the close, the reconciliation, the report better than any developer ever will. Claude handles the syntax. You handle the problem definition.

What it is: Claude writing, running, and debugging code on your behalf. You describe what you need in plain language. Claude builds it.

When to use it: building things that don’t exist yet. A close checklist that lives in a system instead of someone’s head. A dashboard that pulls your key metrics together in one place. An automation that flags reconciling items over a set threshold and surfaces them without you having to hunt for them. A tool that reformats a messy data export into a clean, board-ready table in seconds instead of an hour.

Here’s what a Code prompt looks like in practice:

You didn’t write a single line of code. You described the problem. Claude built the solution.

The thing most accountants don’t realize: the first build doesn’t have to be perfect. It doesn’t have to be scalable on day one. It just has to work better than what you had before. Start with one process that frustrates you every month. Describe it to Claude. See what it builds. Refine from there.

Bringing it all together

Three modes. One decision framework.

Ask yourself: what am I actually trying to do?

If you need to draft, summarize, explain, or refine something — Chat.

If you need to hand off a complex multi-step task and get a complete output back — Cowork.

If you need to build something that doesn’t exist yet — a tool, an automation, a dashboard — Code.

They’re not mutually exclusive either. The best workflows use all three. You might Chat to define the problem, use Code to build the automation, and then hand the output to Cowork to summarize and package for the CFO.

That’s the full toolkit. That’s what it looks like when an accountant stops using 10% of what’s available to them.

Pick one task from your current workflow and deliberately use a mode you haven’t tried before. If you’ve only ever used Chat — describe a process to Claude and ask it to build you something. If you’ve never tried Cowork — hand it a multi-step task and see what comes back.

The gap between where most accountants are and where they could be isn’t talent. It isn’t access. It’s awareness.

Keep building.

Here’s my question to you:

Which of the three modes are you most excited to try — and what’s the first task you’re going to use it on? Drop it in the comments — I read every one.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

Not the fundamentals — debits still equal credits, the close still needs to close, the audit still needs support. Those aren't going anywhere. The foundation is solid and it's earned its place.

But the playbook has always carried its own weight too. Over-reliance on human action. The time it takes to close the books. Systems that don't connect the way they should. Accountants who can't get out of the back office long enough to have a voice in the room that matters.

What's changing is everything built on top of that foundation. The tools. The workflows. The role itself. And for the accountants paying attention — the identity.

That's what this article is about.

What changed

The tools have arrived.

AI, automations, vibe coding and agents — the kind of tools that used to require an enterprise budget and a dedicated IT team are now accessible to team members for less than the cost of a monthly streaming subscription and are now part of the nomenclature of the profession. The barrier to build has essentially disappeared.

However, the profession hasn’t caught up.

The perception problem runs deep. Accounting has always been the back office. The cost center. The department that closes the books and hands the numbers upstairs. That identity was built over decades and it doesn't change overnight.

But here’s what’s shifting: the accountants who have discovered what these tools can do aren’t waiting for the curriculum to catch up. They’re on the front lines setting the standard for how these tools should be used. They’re becoming builders. Quietly, in the background, producing work in hours that used to take days — and starting to ask a question the profession hasn't had to answer before.

What happens when one person can do what used to take a team?


The Accounting Engineer

An accountant, by definition, is a person whose job is to keep or inspect financial accounts. An engineer, by definition, is a person who designs, builds, or maintains engines, machines, or structures.

Now put them together.

The Accounting Engineer is someone who has mastered the fundamentals of accounting — the debits, the credits, the close, the audit — and then uses that knowledge to design and build the systems around them.

In the beginning, accounting was entirely human. Every entry, every reconciliation, every report — done by hand, by people, line by line. Then came the spreadsheet. Then came the ERP. Suddenly the work became half human, half automated. The accountant still did the work but the tools did more of the heavy lifting.

We’re now entering the current phase.

The Accounting Engineer builds systems so robust — so well designed — that the routine work runs itself. The close doesn’t take days because the close has been automated. The reconciliations don’t pile up because the reconciliations have been engineered away. What’s left for the human? Review. Judgment. The things that actually require an accountant.

That’s the point. Not to replace the accountant. But to enable and free them.

The Accounting Engineer isn’t the person doing all the work. They’re the person who built the system so the work gets done — and then walks out of the back office and into the room where the real decisions are being made.

That’s the identity shift. That’s what this role represents.

And it’s built on three things in this exact order: knowing accounting deeply, building the systems confidently, and then letting what you’ve built do the work — so you don’t have to.

An accountant knows the numbers. An engineer builds the systems. The Accounting Engineer does both — and that changes everything.

What this means for smaller companies

At a smaller company, the Accounting Engineer isn’t a title on an org chart. It’s a skill.

Think about what one person carrying the entire finance function actually looks like. Controller. AP. AR. Close. Reporting. All of it. One person, one set of hands, one calendar that never has enough days in the month.

The traditional approach to that problem is to hire. Add headcount. Split the work. But that costs money a small company often doesn’t have — and it doesn’t solve the underlying problem. It just distributes it.

The Accounting Engineer solves it differently.

Instead of adding people, they add infrastructure. They automate the AP workflow. They build the close checklist into a system that runs itself. They connect the tools so the data flows without anyone touching it. They engineer away the routine so the close that used to take ten days takes three.

And then something shifts.

That one person — who used to be buried in transactions — now has time. Time to analyze. Time to advise. Time to sit in the room where the business decisions are being made and actually contribute to them.

At a small company, one Accounting Engineer doesn’t just change the finance function. They change what finance means to the business.

What this means for larger companies

At a larger company, the Accounting Engineer plays a different but equally critical role.

The problem at scale isn’t capacity — it’s complexity. More entities. More systems. More data moving between more places. More people doing more manual work to compensate for infrastructure that was never properly built.

The Accounting Engineer is the person who fixes that.

They sit at the intersection of finance and technology — translating between what the CFO needs and what the systems can do. They’re not just automating tasks. They’re redesigning the workflows those tasks live inside. They’re building the agents that handle the routine, the dashboards that surface the signal, and the infrastructure that makes the entire team more effective.

And they become the architect the CFO relies on.

Not for the numbers — the team handles the numbers. For the systems behind the numbers. For the answer to “why does this still take us three days” and “how do we make sure this never falls through the cracks again.”

At a larger company, the Accounting Engineer is the person who makes the finance function scale — not by adding headcount, but by building infrastructure smart enough that headcount becomes a choice rather than a necessity.

The evolution of the role

A week ago the term "Accounting Engineer" didn't have a home. It was a feeling more than a title — something the most forward-thinking people in the profession were moving toward without having a name for it.

Since then the market has responded faster than expected. At Rillet Recon in New York, the conversation centered on the Finance Engineer — building processes, systems, and agents end-to-end with human judgment at the center. Numeric and Ramp are hosting a webinar called "The Finance Engineers." Not "AI for Finance." Not "The Future of Accounting." The Finance Engineers — a title that would have meant nothing five years ago and today fills a room.

The market is naming the role before the job postings exist. That's how you know something real is happening.

And it's not just the events. It's the conversations happening inside accounting teams right now. The controller who automated the close and suddenly has ten hours back a month. The senior accountant who built a dashboard their CFO now opens every morning. The staff accountant who vibe coded a reconciliation workflow and cut a three-day process to three hours.

None of them have the title yet. But they’re doing the work.

The Accounting Engineer is emerging from the ground up — not from a job description handed down from a hiring manager, but from accountants who decided to pick up the tools and build something. The title will catch up. It always does.

How to become one

Here’s the thing about the Accounting Engineer — there’s no certification for it. No degree program. No exam to pass or credential to earn.

It’s a decision.

And it starts smaller than you think.

You don’t become an Accounting Engineer by overhauling your entire close process overnight. You don’t need to know how to code, have a background in IT, or wait for someone to tell you what to build. You need one process, one problem, and the willingness to try.

Pick something you do every week that frustrates you. A reconciliation that takes too long. A report you rebuild from scratch every month. A close checklist that lives in someone’s head instead of a system. Something small enough to tackle but real enough to matter.

Then build a solution.

Use the tools that are available to you right now. Prompt an LLM to help you think through the workflow. Vibe code a simple automation. Connect two systems that should have been connected years ago. It doesn’t have to be perfect. It doesn’t have to be scalable on day one. It just has to work better than what you had before.

That’s the first build. And the first build changes everything.

Because once you’ve done it once — once you’ve seen what’s possible when you stop accepting the process as it is and start engineering something better — you can’t go back. The way you look at every manual task, every repetitive workflow, every “that’s just how we’ve always done it” moment shifts permanently.

That’s the mindset of the Accounting Engineer. Not “how do I get this done?” but “how do I build something so this gets done without me?”

The fundamentals don’t change. Debits still equal credits. The close still needs to close. The audit still needs support. But the Accounting Engineer has built the infrastructure that handles all of it — and then walked out of the back office and into the conversation.

That’s the goal. That’s the identity. That’s what this profession is building toward.

You already know the numbers. Now go build.

Here’s my question to you:

What’s the first thing you’re going to build? Drop it in the comments — I read every one.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

Not because the profession is going away. It isn’t. But the version of it most of us were trained for — the head down, follow the process, don’t ask why version — that one is gone. And honestly, good riddance.

I spent my first year at EY doing what I was told. Work hard, follow the path, keep your head down and good things will come. That’s the advice you get early. It’s not bad advice. For a lot of people it works. But at some point I started noticing that the accountants moving fastest weren’t necessarily the most technical. They weren’t the ones who knew every standard or had the cleanest workpapers. They were the ones who looked at a broken process and asked — why are we still doing it this way?

That question is the whole thing.

The New Age Accountant isn’t a job title. It’s a way of operating:

• Open to change

• Willing to break something that technically works if it could work a lot better

• Not paralyzed by the fact that AI is reshaping the profession faster than most people feel comfortable admitting.

Being a new age accountant doesn’t mean you’re not scared. Most people I’d put in that category feel the uncertainty too. They just don’t let it create inaction. That’s the real distinction. Not confidence. Not expertise. Just the decision to move anyway.

The mindset shift that changed things for me was getting okay with going against the room. If everyone in the profession says X, that’s worth pausing on. Not because the majority is always wrong. But because the majority is optimizing for what already works, not for what could work. If you only ever ask the questions the room is already asking, you’ll only ever get the answers the room already has.

Most accountants were trained to minimize risk. Flag the error. Follow the standard. Don’t deviate. That’s important in a lot of contexts. But that same instinct, applied to your career and your processes, keeps you doing things the way they’ve always been done. The spreadsheet that takes four hours every month. The close process nobody has touched in three years. The report that goes to leadership that nobody is sure anyone actually reads.

Nobody fixed them because nobody stopped to question them.

The accountants building something right now aren’t doing it because they’re more technical. They’re doing it because they got comfortable being wrong in front of themselves. They tried something, it didn’t work, they tried again. The tools are learnable. The willingness to look uncertain for a minute while you figure it out — that’s the harder part. That’s also the part nobody can hand you.

You don’t have to overhaul everything at once. That’s not what this is about.

This Week’s Action: Pick one routine task you do regularly. Something straightforward, something you’ve done so many times you don’t really think about it anymore. Take it to an AI tool and ask one question: how could I do this better or automate it?

Don’t overthink the prompt. Just ask. Then pressure test what comes back. Does it actually work? What would need to change? You’re not committing to anything. You’re just starting to look at familiar things differently.

Accounting as we knew it is dead. What you build next is up to you.

Quick recap: Vol. 1 covered AI & LLMs — what they are, which ones exist, and how to choose one. Vol. 2 introduced prompting and the CO-STAR framework — the foundation of getting outputs that are actually usable. If you missed either, go back and start there (using links provided) as this series builds week by week.

Now, before we go any further — did you use the CO-STAR framework to enhance your prompting? Because if you did, one of two things likely happened. Either it worked better than expected, and you’re starting to see what all the hype around AI is. Or, it worked okay, the output was close, but you had to rewrite more than anticipated.

If it’s the latter, that’s perfect — that’s exactly what Vol. 3 is for.

Why good prompts still fall short.

Here’s something I’ve noticed working with these tools daily: even a well-written prompt can fall short if the LLM doesn’t have the right frame for the task.

What does that even mean?

Think about how you’d provide context to a first year associate or new accountant on the team. You wouldn’t just hand them a task and walk away. You’d walk them through their specific role in the process, what standard you’d like set, what a great outcome looks like, and potentially provide examples of something in the past (SALY, if you know, you know).

That context — or the frame around the task — is really what separates a good output from a great one.

Good outputs come from people giving the LLM the task. Great outputs come from giving the LLM both the task and the frame.

Let’s get into the meat of it. Here are four techniques to creating a strong frame to immediately improve your outputs. None of them are complex but all of them provide real value to your prompting skills.

Technique 1 : Establish a role

This is going to sound strange. But, before you describe a task, tell the LLM who it is. Weird, I know. However, when you set the stage for the LLM, it activates a specific body of knowledge and strives to achieve in line with that specific standard.

Here’s an example without a role:

Here’s an example with a role:

Same task. However, the second prompt will produce an output that is materially more useful because the LLM is approaching it from a specific lens, instead of a general one. This works for controllers, auditors, analysts, and more.

Try it on your next prompt. The difference will be immediate.

Technique 2 : Provide an example

Just like a first year or someone new to the company, if you want a specific output format, provide an example to the LLM.

Instead of describing the format you’d like to see — which is much more difficult than it sounds — just provide an example of something similar for it to work from. It’s as simple as uploading the file and saying, “use the uploaded document as a template”.

Here’s what that looks like in practice. Say you want the LLM to write a board-ready variance commentary. Instead of writing “format it like a board report with bullet points and an executive summary” — just upload last month’s commentary and say “match this format exactly.“ The LLM picks up the tone, the structure, the level of detail, and the language all at once.

Once provided, the LLM then has a template, tone, task and more to work from. The output will match your format far more closely than if you tried to describe it from scratch.

An important thing to note, once you’ve created this task, make sure you save the prompt with the example baked in. I typically rename my chats and pin them, if I’m going to be using them later on. If you do this, you will be able to use the same chat every month and it saves more time and tokens because you aren’t reprocessing the prompt end to end.

Technique 3 : Ask to be asked

Hands down, this is one of the most underused techniques. Before you write a complex prompt, try this:

Simple, right?

Two things happen when you do this.

First, the LLM finds gaps and weak points in your prompt you likely didn’t know were there. It’ll ask about your audience, your format preference, the context behind the numbers, the tone, and more. It takes the CO-STAR framework to the next level. Most of which, are things you would have forgotten to include.

Second, it teaches you how to prompt better over time. The questions AI asks you are the exact things you should be thinking about and including in your prompts by default. After a few rounds of this, you’ll start anticipating these questions automatically.

Any time I’m prompting something new for the first time, I do this. It almost always produces a better first output than going straight to the task.

Technique 4 : Refine, don’t restart

This one is a mindset shift as much as a technique.

When the output isn’t quite right, most people start over. New prompt, from scratch, trying to get it right in one shot.

Don’t do that.

Remember what Vol. 1 said: you’re not Googling. You’re having a conversation. That means you can — and should — push back, redirect, and refine within the same conversation.

Here’s what that looks like:

First output comes back. It’s okay but too formal for your CFO’s style.

Instead of starting over, say:

Each refinement builds on the last. By the third or fourth exchange, you have something significantly better than anything a single prompt would have produced.

The conversation is the tool. Use it.

Bringing it all together

Alright. We’ve gone through a lot in the last few weeks. Here’s what a fully loaded prompt looks like using everything from Vol. 2 and Vol. 3 combined:

That prompt has a role, an example, real context, and an instruction to clarify before it starts. That’s about as strong as a prompt gets.

Try it this week on something from your actual close. See what comes back. Refine once. Save it.

The accountants putting these into practice right now are producing work that would have taken hours in a fraction of the time — and they’re getting better at it every week. That’s the compounding effect of learning this properly. Vol. 4 is next. Keep building.

Here are my questions to you:

Which of these four techniques are you most excited to try — and which one surprised you most? Click below and tell me. I read every response.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

Then I left.

Not because I had a plan. Not because something better was lined up. My gut said go, my brain said stay, and I went anyway. Everything I’d worked toward, the firm, the path, the thing you’re supposed to do, I walked away from it.

That moment didn’t feel brave at the time. It felt like a mistake I was making in slow motion.

I spent the next few years doing things that had nothing to do with accounting. Customer success. Technical implementations. Cross-functional projects, client calls, broken processes I had to fix with people I’d never met. It felt like a detour. It probably looked like one too.

I didn’t see what I was actually building until I was past it.

Today I’m a controller at a Series B fintech startup. My dream job. I wouldn’t be here without the detour. The communication skills, the project management instincts, the ability to work across teams, none of that came from accounting. It came from the years that didn’t look like a career.

That’s what Beyond the Books is about.

Everyone is talking about AI right now. The tools, the automation, the workflows. Brock covers that well in AI for Accountants and The Builder’s Playbook. But nobody is talking about the person who has to actually use it. The mindset that makes someone ready to pick up a new tool and do something with it. The habits and attitudes that separate the accountants who adapt from the ones who freeze.

AI is going to reward a specific kind of person. Not the most technical one. Not the one who knows the most. The one who is curious, adaptable, and willing to move before they feel ready. Those aren’t technical skills. They’re career skills. Life skills. And they’re learnable.

That’s the gap this fills.

There’s a version of you reading this right now that fits one of these:

You’re a year or two into a big firm and quietly wondering if this is it. You’re a manager who feels behind and doesn’t know where to start. You’re a new grad trying to figure out which direction to even run. You’re mid-career and watching the profession change faster than feels comfortable.

I’m not going to tell you the path is linear. For me it wasn’t. I’m not going to tell you traditional advice is wrong either, for a lot of people it works. But it isn’t the only way, and not enough people say that clearly.

Every issue I’ll give you one thing you can actually do. Not a mindset shift. Not a vague suggestion. Something real.

This week: Write down three skills you’ve built outside of pure accounting work. Jobs, side projects, life experience, anything counts. Then write one sentence next to each one explaining how it makes you better at what you do today. If you can’t connect them yet, keep the list anyway. You will.

The path doesn’t always make sense while you’re on it.

That’s okay. Stick around.

Accounting is at an inflection point. The tools are available — accountants are already automating close processes, building dashboards, and writing code without a CS background. The main hurdle? Accountants who are willing to pick them up and build. Builders, not bookkeepers. That’s what this is about.

Here’s how it works: each issue we will walk through a real workflow, automation, vibe coding, or build — all from real accountants. We will highlight the problem, the steps, and the outcome. Whether the use case is applicable to you or not, the hope is it proves that this is real — and that you are more than capable of becoming a builder yourself.

Have you built something — or at least tried to? This newsletter is backed by real stories from real accountants. Not just us writing but actual operators with unique challenges and perspectives. If you’ve automated a process, built a workflow, or even just started experimenting, I want to hear from you! Hit reply and let me know what you’re building.

The future of accounting has a lot of uncertainty. However, I’m certain the profession will be full of builders — systems architects, automators, and creators. Maybe even Accounting Engineers. Time will tell but I’m very bullish on the future of accounting — and even more bullish on the accountants who are already building it.

Here’s my question to you:

What is something you have built or want to build? Tell me below and let’s try to build it. I read and respond to every comment.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

1/ More often than not, the issue isn’t the tool — it’s the prompt.

2/ If the only AI you’ve used is the chat feature, you haven’t even scratched the surface. But we’ll save that for another day!

The art of prompting

What is prompting? Simple answer: the questions, comments, and/or instructions you provide to an LLM. It can be one word or a paragraph, a question or a command, simple or detailed. Whatever you put into the little box (see image below) is considered a prompt.

General tip: the more specific the prompt, the more effective the output. For example, if you type “help me with this reconciliation” — but fail to add additional context — the LLM has no idea what reconciliation you’re referencing, what account, what steps you’ve already taken, the format of the output, etc. It will provide something very generic because that is all it has to work with.

The fix is simple: be more specific. So what does ‘more specific’ actually look like? That’s where CO-STAR comes in. There are several prompting frameworks out there. CO-STAR is the one I keep coming back to, and the one I believe translates best to accounting workflows.

Reference to image.

Context: The background information related to the situation, project or topic. Example: “We are a mid-sized manufacturing company closing out the Q4 books.”

Objective: The specific goal or task you want to be accomplished. Example: “Reconcile accounts payable and flag any discrepancies over $500.”

Style: The writing style you want. Example: “Present this in a formal, financial reporting style.”

Tone: The emotional feel of the output. Example: “Keep the tone professional, conservative, and audit-ready.”

Audience: Who the output is for. Example: “This report is intended for the CFO and the external auditors.”

Response: The format and length you expect. Example: “Provide the journal entry as a CSV file.”

Here’s what that looks like in practice. An example of a weak prompt would be:

An example of a strong prompt would be:

Same desired output, completely different results. The first may produce a template or a rough outline. The second gets you something you could actually send to the intended audience. The more specific, the better.

Start with one prompt this week. Use CO-STAR. See what comes back. Next issue, we’ll dive deeper into the trenches. But prompting is the foundation — get it right first.

Here’s my question to you:

What’s a task you do every week that you haven’t tried prompting yet? Why not? Click the button below and tell me. I read every response.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

Whether you are just starting out or are a seasoned veteran, one of the first ways to get into AI within accounting is through an LLM (large language model). An LLM is simply a type of AI that is trained to read, write, and reason using language. The most notable are Claude, ChatGPT, Copilot, and Gemini. If you’ve ever typed a question into one of those, you’ve already used one.

Each of these models has their own strengths and weaknesses. Claude, made by Anthropic, is known for its strong writing, deep analysis, and complex reasoning. ChatGPT, made by OpenAI, is widely recognizable in the space and is known for its broad capabilities, ease of use, and large user community.

As an example, I once dropped the same request into both Claude and ChatGPT: “Summarize the key risks in this contract.“ Claude returned a structured breakdown with specific flags; ChatGPT gave a cleaner, more readable summary but lacked depth. Neither was wrong. They just approached the response differently. Try the same thing for yourself. Same prompt, two tools. You’ll learn more in five minutes than reading any comparison article.

One thing worth understanding early on is that these tools are not search engines. You’re not Googling a keyword and scanning results. You’re having a conversation. The better you become at explaining what you need (prompting), the better the output.

Here’s what that looks like in practice:

• Weak prompt: “Help me with this reconciliation.”

• Strong prompt: “I’m reconciling our credit card statement to the GL for October. There are 14 unmatched transactions under $500. What are the most common reasons these mismatches happen? Where should I check first and how should I approach this task?”

Same task, completely different results. The more specific, conversational, and detailed you are, the better. The first one is vague. The second provides context — and context is everything.

Once you’ve picked a tool, here’s what to expect on pricing. Each of these LLMs has various tiers beginning at $0 (with limited usage) and scales to $200+ per month (depending on usage and other factors). It’s worth noting that most of the features that will impact your day-to-day only begin to take shape in paid versions.

Like most people, I personally started out using ChatGPT and I still find it to be a great tool. But over time, Claude has become my default of choice. Like many accountants I’ve spoken with, what won me over was the quality of analysis and its ability to handle dense, complex work — which we do daily. This has become more apparent as I’ve moved to higher tiers.

All things considered, when it comes to which is best, it depends on what you’re trying to accomplish and who you ask. There are many opinions. Focus on what you’re trying to accomplish, pick the one you prefer, and get after it.

If you don’t know where to start, here’s a prompt to try (replace words as needed):

I’m an accountant who wants to start using AI in my day-to-day work, but I’m not sure where to begin. Can you give me a simple, beginner-friendly roadmap? Start with the most practical first steps — things I can try today — and explain how AI can help with common accounting tasks like reconciliations, reporting, data analysis, or client communication.

Believe it or not, that is how it started for me. I knew this was the future, it was inevitable, however, I had no idea how to even begin. Now, through countless hours of trial and error, I still find myself asking Claude how to get more out of these tools.

Don’t overthink the prompt either — it doesn’t need to be perfect. One of the biggest misconceptions is that you need to know exactly what to say to get value out of these tools. You don’t. Just start prompting, see the result, and refine from there.

Start small. Keep refining. Sooner or later it’ll make sense and you’ll surprise yourself with what you’re capable of.

The accountants who will thrive in the next decade won’t necessarily be the ones who know the most technical details about how AI works behind the scenes. The accountants I’ve seen get the most out of these tools are the ones who identified a bottleneck, opened a tool, tried to build the solution, and kept going. Stay curious. Keep learning.

It’s worth noting that the same tool a beginner uses to draft an email to a client is the same one advanced users are connecting to live data, automating workflows, and running full flux analysis and scenario forecasting. That’s where this series goes. But it starts here.

Now that you know what these tools are, the next question is: how do you actually talk to them?

That’s where most people get stuck — and that’s exactly what we’re covering next.

Here are my questions to you:

Which LLM do you currently use? Do you have a strong preference to a specific LLM? Click the button below and tell me. I read every response.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you’re a staff accountant, a controller, or a CFO, there’s something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you’ve made it this far, you’re already thinking differently about this profession. Subscribe and come build with us.

Subscribe now

Founded by the ‘Father of Accounting,’ Luca Pacioli, in the 1490s, it’s been the universal language through which the world does business — beginning with the abacus, adopting the ledger, evolving to spreadsheets, and arriving at the present day.

By nature, accountants are creatures of habit. Routine is essential — and for good reason. Consistency, accuracy, and repeatability are the backbone of timely and accurate financial reporting.

For years, the playbook has been the same:

1/ start with QuickBooks until you’re ready for NetSuite,

2/ perform manual work for heavy transaction volume (keeping you in the back office),

3/ hire, hire, hire,

4/ repeat.

For years it has worked; I’ve seen this firsthand — both from an audit lens and from in-house accounting roles. It always left me with one question: Why does it feel like we’re so far behind the times?

Well, the times have changed.

So what’s changed? Simple answer — everything.

Spend management platforms like Ramp and Brex have made expense reporting and corporate cards nearly painless. AI-native ERPs like Rillet are beginning to replace the clunky, legacy systems we’ve tolerated for decades. Sales tax compliance tools like Sphere and Anrok have turned one of accounting’s biggest headaches into something that practically runs itself. And this is just the tip of the iceberg.

Then came January 12, 2026.

Anthropic released Cowork — and the accounting world hasn’t been the same since. Not an exaggeration: it may be the single most transformative tool to ever touch the profession.

And it didn’t stop there. Vibe coding tools like Lovable and Replit have made it possible for accountants — not developers — to build their own custom tools. No coding degree required.

What’s the new playbook? Here’s my take:

1/ implement an AI-native ERP that scales as you scale,

2/ automate routine tasks through AI-enabled software and systems (e.g. expense management, sales tax, etc.),

3/ leverage Cowork and other tools to create dynamic reports, processes, and routines,

4/ become a strategic accountant who isn’t kept in the back office but assists with high-level strategy and cross-departmental growth,

5/ repeat.

The future of accounting isn’t coming. It’s already here. Start taking action now — break away from the creature of habit stigma of the traditional accountant. It’s time to become builders, not bookkeepers.

Here’s my question to you:

What are you the most excited about? Click the button below and tell me. I read every response.

Leave a comment

The purpose of New Age Accounting is simple: to empower accountants — at every level — to become builders, not bookkeepers. Whether you're a staff accountant, a controller, or a CFO, there's something here for you. Some topics will be high level, others will come with step-by-step guides, and some will include the exact prompts and tools you need to start building today.

If you've made it this far, you're already thinking differently about this profession. Subscribe and come build with us.

Subscribe now