> ## Content Index > Fetch the complete content index at: https://www.financealliance.io/llms.txt > Use this file to discover other available public pages before exploring further. # Stop fearing AI and start building with it: A finance leader's view from the trenches - URL: https://www.financealliance.io/stop-fearing-ai-and-start-building-with-it-a-finance-leaders-view-from-the-trenches/ - Published: 2026-10-12T07:00:13.000Z - Updated: 2026-10-12T07:00:13.000Z - Description: Doing nothing is a sure way to make our functions and organizations obsolete. We need to jump ahead and become the accelerator and the enabler of the business. - Author: Lydia Stone - Tags: AI & Automation, Articles I started my career as a traditional accountant at Big Four firms, spending my first 10 years with large public enterprises. That gave me a huge learning ground for rigor, discipline, [technology](https://www.financealliance.io/rethinking-finance-teams-for-the-age-of-ai/), and transformation. Eventually, though, I found my home on the tech and growth side, helping companies through their growing pains. At Chargebee, my job has three parts. I'm the guardian of the numbers, responsible for making them, reporting them, and helping people understand them. I use technology to make sure [finance](https://www.financealliance.io/) is an accelerator for the company rather than a bottleneck. And because we eat our own dog food, I also advise our product team on the future of our product. That vantage point has given me a clear view of where [finance teams](https://www.financealliance.io/how-fp-a-teams-build-budgets-and-how-you-can-nail-this-in-interviews/) really are with AI. --- **This article is based on Lydia's** [**brilliant podcast episode**](https://fast.wistia.com/embed/channel/bk71ekafa7?wchannelid=bk71ekafa7&wmediaid=go13e4bnuf)**, which you can watch in full here:** --- ## **Where finance teams really stand** Beyond the hype, most people are still experimenting and piloting. There are successes in data extraction, [data analysis](https://www.financealliance.io/mastering-data-storytelling/), and reporting efficiency, but beyond that I see a strong desire to use AI more, combined with real hesitation. I think of it as a trust gap. Finance people are inherently concerned about accuracy. Can I trust the [process](https://www.financealliance.io/17-finance-business-processes/), and therefore the numbers? I don't believe we're there yet. Most teams have either started piloting, are in the middle of it, or have finished a pilot without knowing what comes next. Only a small minority have figured out a midterm plan. The pace doesn't help. Yesterday I figured out ChatGPT, today it's Claude, and by tonight ChatGPT has released a new version. Which tool is better at which task? We're trying to build [trust](https://www.financealliance.io/when-should-you-actually-trust-ai-in-a-finance-decision/) while the ground keeps shifting. [Omnichannel finance: Navigating the finance frontier of multi-vertical growthIf there’s one thing I’d want people to take away from all of this, it’s that a successful strategy comes down to simplicity and focus.![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/icon/android-chrome-192x192-ebcc566a-4f39-49c9-a25c-8df359aadbe1.png)Finance AllianceFederico Reyes![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/thumbnail/Copy-of-FA_Website_Article_Images_Author_Highlight--44--bfca4c18-b3bf-4a1b-ba1d-e7e107028e21.png)](https://www.financealliance.io/omnichannel-finance/) ## **Solving pain points, not replacing jobs** Most of the value AI brings to finance today doesn't come from automating an entire job. It comes from solving pain points within existing processes. Individual task agents are making our [forecasting](https://www.financealliance.io/top-down-vs-bottom-up-forecasting/), [reporting](https://www.financealliance.io/what-is-a-compliance-report-compliance-reporting-finance/), reconciliations, and [month-end](https://www.financealliance.io/month-end-close-checklist/) close faster by taking away the data digging. AI has also changed my own routine. The first app I open in the morning isn't email anymore; it's Notion. My agents there summarize my calendar, warn me about conflicts, remind me about pre-reads, summarize emails I missed, and flag actions I need to take. I almost take it for granted now, which shows how fast things move. The most exciting realization for me is how powerful my team feels. If we need a small tool to solve a problem, we no longer have to hand business requirements to engineers and wait through coding, SIT, UAT, and a go-live date. We're the process owners. We know how things should work, and in this no-code world, we can build it ourselves. ****Since May 2025, that has felt like having two professions in one, and it feels amazing.** ## **From individual agents to an ecosystem** Because AI cuts so much prep time, it frees us to partner with other teams. The trend has been that [the CFO is becoming the new COO](https://www.financealliance.io/how-cfos-power-business-strategy/), because every part of the organization matters in getting to the numbers. Now we're building agents that orchestrate multiple parts of the company. Our deal desk agent, which is almost ready to go live, connects sales, marketing, and negotiation with the back end: billing, revenue recognition, [ARR](https://www.financealliance.io/understanding-the-rule-of-78s-a-critical-tool-for-recurring-revenue-forecasting/) reporting, and the board's perspective. Early in a deal, we can already see what the back end will look like, and on the back end, we can see how things are shifting at the front end. It took several smaller agents, architected together, to create that view. We're collaborating without realizing we're [collaborating](https://www.financealliance.io/your-guide-to-building-stronger-cross-functional-relationships/), absorbing information every day instead of scheduling another long meeting. Intuitively, I believe this network of agents working together is where the higher value lies. [What strategy needs from FP&A and why it’s still missingStrategy today is deeply embedded in commercial execution, and compensation increasingly reflects that rather than the quality of a deck.![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/icon/android-chrome-192x192-1f1ac926-f2cd-44fc-a704-94f0b07c5935.png)Finance AllianceNoufal Basheer![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/thumbnail/Copy-of-FA_Website_Article_Images_Author_Highlight--37--594bebd6-d5c2-4b60-8f8e-13a39d5cd223.png)](https://www.financealliance.io/what-strategy-needs-from-fp-a/) ## **How our journey started** We began by trying to solve a painful process. We did demos with the idea of buying a tool, but every time I thought we could probably build something ourselves. Chargebee provides just about every AI tool you need and encourages people to experiment. Around June or July of 2025, one of our team members took a painful manual process and showed us he could eliminate about three days of work in about three minutes. We were flabbergasted. We asked him to teach everyone, and an informal competition broke out, with people claiming they could cut out five or six days. Before we knew it, we had a long list of small [tools](https://www.financealliance.io/15-best-fp-a-tools-and-software/) that each made someone's job easier. Then, as we moved into 2026, Claude made no-code building of more complex agents truly possible. ## **Becoming AI-first** Our philosophy is to be an AI-first company. When you have a question, you don't go Google it; you go AI it. That takes support from both a technology and a cultural perspective. When we asked our team to list pain points they wanted solved with AI, the list was very long. We realized our AI engineers couldn't build it all, so we left them the complex projects needing lots of APIs and integrations, and told everyone else to build their own. We're not at full scale yet, but we've moved beyond simple task-level agents, and our first collaborative network of agents is about to go live. [Operationalizing productivity in a $1B+ construction verticalWhen you treat your OpEx budget as a dynamic engine rather than a static bucket, you don’t just save money; you create a resilient competitive advantage that scales alongside your infrastructure.![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/icon/android-chrome-192x192-e342b5b3-49fb-476e-badd-d971dd7aa45e.png)Finance AllianceAlexander Salamandra![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/thumbnail/Copy-of-FA_Website_Article_Images_Author_Highlight--27--280a2260-2962-4c61-8a84-6383de5ad7fd.png)](https://www.financealliance.io/operationalizing-productivity-in-a-1b-construction-vertical/) ## **My advice for getting started** Start with your pain points, start with what you know best, and don't try to be perfect. We had manual spreadsheets we had to stitch together by hand, and we thought AI could do it. It did, eventually, but it made mistakes we didn't expect and took longer than planned. That was a learning process in itself, because you learn which parts you can trust. Nobody has the ultimate answer on how to use AI, so start small and take the first step. ## **Treat AI like a brand-new employee** Don't expect AI to be a magician. It's more like hiring a brand-new staff member. You walk them through your process step by step, and when they get step five wrong, you give them context about how you deal with customers and vendors and ask them to try again. You verify, course correct, and add context. We had a tool read hundreds, if not thousands, of our nonstandard contracts to catch extra clauses and terms that traditionally required a human reader. At first, it hallucinated so much that the error rate was unacceptable. So we literally told it, "Please do not fabricate data. If you're not sure, don't give it to me," just as you'd tell a new hire. Train it like a new staff member on day one, and it can gradually become more senior. [Top 5 communication mistakes to avoid for finance & FP&AThere is no right or wrong way to communicate, but there are clear differentiations between a concise, effective, and impactful communicator versus one that is wordy, robotic, and inconsistent.![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/icon/android-chrome-192x192-c6e600c1-f277-4f57-b44d-a5b46b557fb5.png)Finance AllianceRicky Koo![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/thumbnail/Copy-of-FA_Website_Article_Images_Author_Highlight--40--c32fe2c0-19f2-441a-a96b-02e9a519fc35.png)](https://www.financealliance.io/top-5-communication-mistakes-to-avoid-in-finance-and-fp-a/) ## **Learning the hard limitations** Every tool reads one spreadsheet well, but throw 20 at it and it makes all kinds of mistakes. That gave me serious pause, because the real value lies in stitching those spreadsheets together. I kept correcting it, and it kept repeating the error. If it had been my staff, I'd have started yelling. Eventually I asked why it was struggling, and it pointed me to a video about AI's limits in processing large volumes of Excel data. Useful, but only after I asked. Now I don't upload 20 spreadsheets and expect a result. I've also noticed that the more often ChatGPT says yes to me, the more uncomfortable I get. In finance, critical thinking and seeing the full picture are essential. ## **Building trust in the output** This is why adoption in finance moves at a cautious speed compared to the hype around AI capabilities. It starts with data. No AI tool or human can make a flower out of garbage. Clean up your data, though you don't need to wait for perfection, because AI can help fix it. Then start with lower-risk tasks to build confidence. [When should you actually trust AI in a finance decision?The frame of “trust AI or don’t trust AI” is too simple. The better frame is: is this output ready to be acted on? That requires you to ask hard questions.![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/icon/android-chrome-192x192-1ea93b9a-a440-4dce-9079-8a95fd1cc95d.png)Finance AllianceTeresa Garanhel![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/thumbnail/FA_Website_Article_Images_Doodles--26--9af8b66c-d436-4596-a25f-39e0b24aa5fc.png)](https://www.financealliance.io/when-should-you-actually-trust-ai-in-a-finance-decision/) **If you buy**, choose domain-focused AI vendors that understand your risks, your quality standards, and your [stakeholders](https://www.financealliance.io/stakeholder-communication-plan/), so you don't have to audit their processes yourself. **If you build**, as we do every day, keep a human in the loop, more than you might think, especially in this first phase. ## **Audit the input and the logic** In the past, a fat-fingered number would produce a wacky output, maybe a 3,000 percent growth rate, and you'd work backward to find the error. Now you have to check your inputs and audit the logic of the AI's brain. Ask it to spell its logic out. AI can produce things no human could, like a trend chart covering 15 years of company history in minutes. But what's your benchmark for trusting it? You have nothing to compare it to. What you can be sure of is that your input is good and that you agree with the logic. Verifying enough to put numbers in front of the board or auditors is a learning process too. [Why human skills are the future of financeAs AI transforms finance, it’s not just about tech, it’s about people. Discover why emotional intelligence, curiosity, and leadership are the real future skills that will set you apart.![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/icon/android-chrome-192x192-e9e6d7a9-f3aa-4f95-b52f-94e3d8a00e93.png)Finance AllianceDaniele Martins![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/thumbnail/Copy-of-FA_Website_Article_Images_Author_Highlight--61--81c9a614-1165-456b-8470-819c7c7e6691.png)](https://www.financealliance.io/beyond-automation-why-human-skills-are-the-future-of-finance/) ## **Unpacking the fear** The fear around AI taking jobs is real, but when you unpack it, it's actionable, and action is the best fear buster. The first fear is that skills will become outdated. When the internet arrived, everyone got trained to use it. The answer here is also training, whether through your company or on your own. What can't you learn from [YouTube](https://www.youtube.com/@financealliance7314) these days? The second fear is unclear guidance from leaders. People don't know whether their leaders are jumping in or watching. Leaders need to set a clear vision and tell employees they have a plan or are making one. The third fear, especially in finance, is about accuracy, auditability, and traceability. That means verifying inputs and logic and keeping humans in the loop. AI isn't a magic wand. It takes real work to make AI work for us. [Navigating uncertainty: Why agile FP&A is your competitive edgeMaster change with agile FP&A. Turn uncertainty into opportunity through flexible forecasting, data-driven decisions, and smarter collaboration.![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/icon/android-chrome-192x192-b4fea03d-ceaf-48f5-81e0-e624f7df21da.png)Finance AllianceImane Haouassia![](https://storage.ghost.io/c/a5/73/a5734519-d6ac-4b17-95b8-f7a3866730ab/content/images/thumbnail/Copy-of-FA_Website_Article_Images_Author_Highlight--33--fab869f3-10c0-445e-a2c2-bf291d1e2348.png)](https://www.financealliance.io/navigating-uncertainty-why-agile-fp-a-is-your-competitive-edge/) ## **What doesn't change: accountability** When people ask what needs to change in [governance](https://www.financealliance.io/finance-and-compliance/), I start with what doesn't change: accountability. If you present the wrong numbers to investors, the machine doesn't get the blame. You do. I review my human employees' work differently from my AI employee's work. With people, I know their typical errors, like fat-fingered numbers or misaligned columns. I can see their formula, spot that they used column D instead of F, or ask why a result is 5 percent and learn they pulled last year's number. Through that interaction, you gain comfort in the output. You can't review a machine the same way, because its logic often isn't as visible as a human's formula. With our contract review agent, we tested for three months before reaching an acceptable 2 percent error rate. It took real training and retraining. Keep the human in the loop until that person is comfortable saying, "You can hold me accountable for this." ## **What I don't want to see in five years** What I don't want to see is finance being stopped by fear. AI isn't perfect, but it's useful in many things, and that's good enough. If you have 100 pain points and you solve two or three, that's a start. The tools are evolving, so you'll be able to do better later. Doing nothing is a sure way to make our functions and organizations obsolete. Our profession has so much to gain from [automating](https://www.financealliance.io/fp-a-automation/) away the repetitive manual tasks that give us the reputation of "let's just wait for finance." We need to jump ahead of that and become the accelerator and the enabler of the business.