AI-assisted financial reporting, FP&A analysis, recurring monitoring, commentary, visuals, and workflow automation

Here’s one practical way I use AI to improve financial analysis.

Before sending an analysis to management, I can use AI to help challenge my commentary.

For example, imagine I write:

“Operating expenses were 8% above budget due primarily to higher professional fees and utilities.”

Instead of stopping there, I can ask AI:

“Act as a CFO reviewing this commentary. What questions would you ask before accepting this explanation?”

I might get questions like:

• How much of the 8% variance relates to each category?

• Are the professional fees recurring or one-off?

• What’s driving the increase in utilities?

• Has this happened in previous months?

• Does the variance affect the full-year forecast?

Now I have a checklist.

I can go back to the actual financial data, investigate those questions, and strengthen my analysis.

The final commentary might become:

“Operating expenses were 8% above budget, primarily due to one-off professional fees. Utilities also remained above budget for the third consecutive month, indicating a recurring cost pressure that may require an adjustment to the full-year forecast.”

Much more useful.

AI didn’t perform the analysis.

It helped me identify where my analysis needed to go deeper.

That’s one of the ways I think finance professionals can get real value from AI.

Not by outsourcing our judgment.

But by using AI to challenge it.

Try this with your next variance commentary:

“Act as a CFO reviewing this analysis. What questions haven’t I answered?”

You might be surprised by what you missed.

#AIInFinance #Accounting #Finance #ArtificialIntelligence #FinancialAnalysis #FPandA #FinanceLeadership #FutureOfFinance

I was halfway through “Will AI Take My Finance Job?” when I made the classic AI-thought-leader mistake:
I stopped talking about AI long enough to build something. 😂

This followed enough conferences to develop a permanent lanyard tan—and hear “single source of truth” from nineteen vendors, each selling a different truth.

Finance exposed the real gap:

Corporate America does not need another dashboard.
We already have enough dashboards to tile a small bathroom and still have leftovers for the kitchen.

Dashboards are corporate gym memberships. We buy them with optimism, open them twice in January, then ask an analyst why the numbers look angry.

And every report is named “Executive Overview,” because the executive still needs another human to provide the overview.

So I took an unexpected detour:

Can a BI dashboard read itself?

Not just display revenue in twelve shades of blue. Actually notice what changed, explain why it matters, identify likely drivers, show evidence and suggest what to do next.

That became aibi—“Abby.”

The naming committee was me, developers who would have called it InsightService_v2_FINAL, and an old “Dear Abby” pun.

“Dear Abby, CPA fell this week. Should I celebrate—or audit the attribution model before Marketing buys a yacht?” 😂

Click the alert that scares you most. It is Choose Your Own Adventure for executives.

The design brief:

“Take me directly to the fire. Bring root cause, three extinguishers and please don’t make me touch a slicer.”

The greatest side benefit may be the state funeral for:
This_is_really_really_the_FINAL_FINAL_v8_USE_THIS_ONE.xls

You know the file. Don't pretend you don't. lol
You remember it... Dead giveaway is the Forty-seven tabs. Three former owners. Two circular references. One macro nobody will touch. And a pivot table with more authority than the CFO.

That workbook is not a file.
It is the undocumented ERP wearing an Excel costume.

Then came the avatar:
AAA: AI Analyst Avatar.
Add animation and it becomes AAAA.
The next release may come with roadside assistance because enterprise architecture eventually needs towing.

But the avatar is easy. The real value is governed context: trusted definitions, lineage, permissions, confidence, auditability and human review.

Without that, AI-enabled BI is a persuasive intern who misunderstood the spreadsheet while using the word “strategic” six times in the board deck.

The future of BI is AI turning passive reporting into active decision systems:
Observe → detect → explain → hypothesize → recommend → act.

And perhaps FINAL_FINAL_v8 can retire to Florida with one broken link.
What workbook or dashboard secretly runs your organization?

🔁 Reposting to help others learn what it actually takes to deploy AI in the enterprise.—not just demo or talk about it.

Please click to follow me
👉 https://lnkd.in/g9erKdfw


#ArtificialIntelligence #AIAgents #SoftwareEngineering #EngineeringLeadership #CTO #TechnologyLeadership #AIEngineering #FutureOfWork

New QuickBooks features for August 2026: Intuit Intelligence automates your books and surfaces important insights, while updates to everyday workflows like bank matching, reconciliation, reports, and bill pay keep them clean.

The AI hype is real. It's also incredibly useful, 𝘪𝘧 you know where to apply it. I think the problem for a lot of people is that the practical examples of how AI can actually help, and how it makes you more efficient, are missing.

Here's my honest take, from a finance seat in a scale-up.
So I thought I'd share some real examples of how I use AI, as the CFO of Orbisk | Zero Food Waste.

In the scale-up phase we're in, capital isn't usually spent on big finance teams. But then all the finance topics still need to be covered:

Monthly reporting, updated rolling forecasts, annual statements, ad hoc analysis whenever the business asks for it.

With a small team, that can become overwhelming.

And it leads to something I want to avoid: only doing the "producing" part of the job, instead of the strategic part, giving guidance and direction to the company.

This is exactly where AI can add value. Because a lot of that producing part can be done with AI.

Some practical examples I've automated, that save me a lot of time:

1️⃣ 𝗠𝗼𝗻𝘁𝗵𝗹𝘆 𝗰𝗼𝗻𝘀𝗼𝗹𝗶𝗱𝗮𝘁𝗶𝗼𝗻 & 𝘁𝗿𝗶𝗮𝗹 𝗕𝗮𝗹𝗮𝗻𝗰𝗲 𝗰𝗼𝗺𝗽𝗮𝗿𝗶𝘀𝗼𝗻
Consolidates the trial balances of all entities and summarises the differences per month, per G/L account and expense group

2️⃣ 𝗣𝗿𝗼𝗽𝗼𝘀𝗲𝗱 𝗮𝗰𝗰𝗿𝘂𝗮𝗹𝘀 𝗯𝗮𝘀𝗲𝗱 𝗼𝗻 𝗹𝗮𝘀𝘁 𝟯 𝗺𝗼𝗻𝘁𝗵𝘀 𝗼𝗳 𝘁𝗿𝗮𝗻𝘀𝗮𝗰𝘁𝗶𝗼𝗻𝘀
Spots which vendor invoices are still missing on day 1 of the month, based on last month's pattern, including the average amount

3️⃣ 𝗥𝗼𝗹𝗹𝗼𝘃𝗲𝗿 𝗼𝗳 𝗺𝘆 𝗳𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝗺𝗼𝗱𝗲𝗹
Automatically drags forward all formulas from the last forecasting month to the actual month at month-end

4️⃣ 𝗕𝗮𝗹𝗮𝗻𝗰𝗲 𝘀𝗵𝗲𝗲𝘁 & 𝗰𝗮𝘀𝗵𝗳𝗹𝗼𝘄 𝘀𝘁𝗮𝘁𝗲𝗺𝗲𝗻𝘁 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻
Based on the trial balances in our ERP, it builds the consolidated balance and cashflow statement automatically, including management corrections and a forecast based on the drivers we've discussed

5️⃣ 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗰, 𝗶𝗻𝘁𝗲𝗿𝗮𝗰𝘁𝗶𝘃𝗲 𝗛𝗧𝗠𝗟 𝗺𝗼𝗻𝘁𝗵𝗹𝘆 𝗿𝗲𝗽𝗼𝗿𝘁𝗶𝗻𝗴 𝗽𝗮𝗰𝗸𝗮𝗴𝗲
Built from an auto-generated input Excel, it creates a dashboard and in- and external reporting where I only manually add comments and observations

All of these used to be repetitive, boring, producing work.

I already hear some finance people say: nice story, but are the numbers correct?

Yes. If you build in testing, let AI crosscheck what it produces, and make sure it's always tied back to your source data.

The only catch: you have to invest the time to make it to work. Spend one hour on it and it won't work. Put in the time properly, and it pays you back many times over.

↳ Where in your finance function is AI actually saving you real hours?

Here's the full Claude system I use for finance work.

(Everything a finance team needs to stop wasting hours on manual work)

👉 Comment "Claude" and I'll send you the full set.

I'm limiting this to the first 1,000 comments.

Businesses bring me in to run finance workshops and show their teams how to actually use AI for real finance work.

Most teams I train are using maybe 5% of what Claude can do.

They watch a few tutorials. They paste in random prompts. They get mediocre results and conclude it's overhyped.

No one ever gave them a structured way to use it for the actual job: the reporting, the analysis, the board pack.

So I packaged my best resources into one set: The Claude Finance Toolkit.

Inside, you'll find:

1. The Free Claude Course for Finance
2. Top 50 Claude Tips
3. Top 50 Claude Shortcuts
4. The Claude Finance Playbook
5. Claude for Board Packs & Presentations Guide

Practical guides, prompts, and workflows to save time, sharpen your numbers, and make better decisions.

I've watched owners burn an entire weekend on a board deck that Claude could draft in 30 minutes.

This is the shortcut I wish every finance team had on day one.

👉 Comment "Claude" and I'll send you the full set.

♻️ Repost to help a finance colleague still doing it the hard way.

I built an AI system that replaces your CFO.
Drop a comment and I'll send it to you.
...
I'm kidding but read on and I’ll tell you why.

I’m sure you've seen that post. Probably multiple times this week.

I use AI every single day in my CFO work. My prompts from two months ago look nothing like what I'm using today. The tools are changing fast and the way I use them has to change with it.

Here's what those downloadable systems actually do. They generate reports. They summarize data. They flag obvious variances.

And then they go stale.

There's also a more fundamental problem.

“Crap data in is crap data out.”

If your books are messy and your expenses are miscategorized, your AI CFO system isn't giving you the right insights. It's giving you confident-sounding slightly wrong answers.

These systems work. I use them.

But relying on a static template won't produce consistent results.

Using AI well in financial analysis is an ongoing practice. It requires clean data, good judgment, and someone who keeps evolving it as the tools and the business change.

AI makes a good analyst better. It doesn't make a downloaded template a CFO.

Trintech Flux agent automates account fluctuation analysis during the close; Trintech Variance Analysis agent accelerates budget-to-actual performance review after the close — both operating within governed financial workflows

The more I use AI in finance, the more I find myself thinking less about the technology and more about the people.

A year ago, most of the conversation was about what AI could do.

Could it write commentary?
Could it analyze data?
Could it automate parts of our workflow?

Today, I think the more interesting questions are different.

How do junior professionals develop business judgment if fewer tasks start from a blank page?

How do we make sure people still understand the "why" behind the numbers instead of just reviewing the output?

And what skills become more valuable when everyone has access to the same tools?

I don't have all the answers.

But the more I experiment with AI in my own FP&A work, the more convinced I am that the future of finance isn't really about the technology.

It's about how we adapt around it.

Using AI in accounting feels like working with Einstein — except you have to constantly teach Einstein what accounting is.

A little over 2 months into an AI startup as a Controller, I am having more fun than ever automating accounting workflows so I don’t have to do the work myself. I blew through my session limits 3 days in a row this week (consecutively mind you - 11am to 6pm, 6pm to 11am, rinse and repeat) - but I created financial reporting packs in days that would have taken an entire team weeks to prepare and review.

I no longer want to prepare any workpapers in Excel - that’s my agent’s job now. I am only going to review my agent’s work. Anything in Excel can be automated - no exceptions. In fact, this probably applies to the entire Office suite.

With that said, I am learning that the bottleneck is becoming… me. We have gone full circle - I remember back in my audit days waiting for CaseWear IDEA to finish selecting samples so I can begin testing, or waiting for Blackline to batch process reconciliations so I can start preparing reports. But now, I find myself struggling to keep up with all the different sessions and agents that are running in my Claude. Jumping between sessions to review AI output is harder than expected because of constant iterations. You add jumping in and out of meetings on top and it very quickly becomes absolute ADHD level chaos.

Because the AI works SO FAST, you need to speed up your review frequency as well, and constantly iterate to achieve the desired output. Gone are the leave-a-comment-and-wait-for-your-team review cycles - AI does 10 cycles in 1 minute.

Am I the only one hitting this attention span wall?

Claude’s new financial services release shows how powerful AI in finance is becoming. It also highlights what finance teams still need around it.
The ability to analyze data, work across documents and spreadsheets, draft outputs, and support complex finance workflows is genuinely impressive. For small businesses, solo operators, and lean finance teams, this can already be a major unlock.

But when we speak with finance teams in larger and growing companies, we hear a consistent pattern: they are excited about AI, but the real need is not just better analysis. It is getting finance work done safely, reliably, and with the right controls.

Finance teams tell us that the hard part is often connecting the right context: accounting data, approvals, budget ownership, CRM inputs, tax rules, entity-specific logics, and manager decisions. They also want AI outputs they can inspect, verify, and approve before anything consequential happens.
That distinction matters. AI can prepare, check, summarize, and recommend. But when the next step affects a payment, reimbursement, approval, accounting treatment, audit record, or ERP write-back, companies need more than a smart assistant. They need a governed finance workflow layer.

That is where the next phase of AI in finance will be won: connected data, encoded business logic, permissions, approval flows, audit trails, and human control over consequential actions.
Finance teams should start now. The best place to begin is with workflows where AI can safely prepare the work, while the company builds the operating layer needed to execute with confidence.

The companies that benefit most will be the ones that start learning now. Not by handing over consequential finance decisions on day one, but by putting AI into real workflows, understanding where it creates leverage, and building the controls needed to scale it safely.

How do you think about this? Would love to hear more thoughts

🚀 And check out how Moss can support you on that journey! 🚀