How we’re scaling a lean finance team with AI — and what’s next

Imagine the CEO told you the company plans to double revenue next year and came to you for a plan on how finance can support that. You would likely default to hiring more accountants and analysts to support the increased workload.

Now imagine the same CEO asks you to increase revenue by 10x next year. You would quickly realize that strategy isn’t feasible. That you need to go back to the drawing board and attack the challenge in new, creative ways.

That’s the message behind a book that the leadership team at my company, Packer Fastener, recently read: 10x Is Easier Than 2x. The book reflects how I think about scaling the finance function as our business, a national distributor of nuts, bolts, and other industrial supplies, is in hypergrowth.

Packer has quadrupled our distribution centers and revenue since I joined just four years ago, so I’ve had to embrace the 10x mindset. It requires constantly experimenting with the latest tech and running with what works (my previous life as an IT leader comes in handy here). While Packer may be in an old-school industry, I think our finance team is doing some pretty cutting-edge stuff.

Packer has quadrupled our distribution centers and revenue since I joined just four years ago, so I’ve had to embrace the 10x mindset.

I’ll share a few strategies we’ve adopted to keep finance headcount flat despite rapid growth, then get into future breakthroughs as new AI capabilities continue to emerge.

What we’ve already automated

• Document processing: It amazes me how quickly tools that can read and analyze documents have improved. Today, my staff simply drops bill PDFs into our AP system and, with the help of AI, it accurately pulls in not just totals and invoice numbers but line item amounts and codes. While we’re still refining some capabilities to our workflows, I expect this will save my accounting staff hours every week and prevent the small mistakes that turn into big headaches come month-end.
• Contract analysis: When I evaluate where I spend more time than necessary, reviewing contracts for risky terms sits at the top of the list. So we’ve started feeding contracts into AI. The tool flags any deviations from our standard terms and suggests changes, cutting review time by a few hours per month while reducing risk.

Comparing insurance provider bills to payroll data used to take my team two days every month. With AI, it takes 10 seconds.

• Insurance reconciliations: Comparing insurance provider bills to payroll data used to take my team two days every month. With AI, it takes 10 seconds. In just a few weeks, a team member created a Microsoft Copilot bot that compares the two data sources and flags any discrepancies. We’re already seeing compounding returns: that same person reallocated the time they saved to build a new automation for enrolling new hires in benefits.
• Extensive self-serve reporting: It’s early days, but our data analytics platform lets employees use AI to “converse” with the data. A department leader can pull up last month’s financials and ask why shipping supplies were higher than plan and above the prior month. An operations manager can ask why COGS has spiked 10% in two months. This means finance has far fewer of these requests to handle and decision-makers get immediate answers for faster, better-informed decisions.

Thanks to these automations, our close time has dropped from 20 days to five — and we recently closed in just three days.

What I see coming

• Automated collections calls: No one on my team wants to make the uncomfortable call to that customer again about a bill that’s 45 days past due. As voice AI matures, I can see AI agents handling a lot of these calls, with agent-to-agent conversations becoming the norm. It could eliminate work no one likes and boost cash flow by ensuring that every past-due invoice gets prompt attention.
• Negotiations on autopilot: In our freight business, we’ve already encountered agents that will call carriers to negotiate rates. It’s not hard to see all sorts of applications for this, like vendor contracts and insurance renewals. If these agents can sound like humans and follow complex decision trees to evaluate offers and respond intelligently, teams might secure better deals that lift the bottom line.

I recently used Replit, an AI coding agent, to build an IRR model for our business in about 10 minutes.

• Advanced models in minutes: Analysts know the grind of building a complex model from scratch. I recently used Replit, an AI coding agent, to build an IRR model for our business in about 10 minutes. I was truly impressed: it turned a brief description into a tool with a beautiful UI and some parameters I hadn’t even considered. Less time on mechanics means more time to focus on interpreting results and guiding strategy.
• AI-powered compliance checks: I don’t think AI will ever fully replace human oversight in financial reporting, given the supreme importance of accurate numbers. But I do see agents taking over much of the prep work. That means automatically reconciling subledgers and checking filings against regulatory requirements. This could shorten close cycles and reduce the risk of compliance issues while letting controllers concentrate on work that requires deep subject matter expertise.

3 easy steps to try today

I get the fear of falling behind, so I suggest starting with these simple habits to build momentum:

You’ll be surprised how far this gets you, and soon you’ll have a finance team that operates more efficiently and is ready to embrace the change that’s inevitably coming.

If you’d like to connect or want to go deeper on anything I covered here, please connect with me on LinkedIn.