
A two-year-old bookkeeping firm. 12 employees. Every single client closed by Day 4.
Here's how they built the system that made it possible.

Aaron Ressel spent his career as a financial controller, while his partner Kevin spent his as a CFO. They were good at their jobs. But everywhere they went, they faced the same challenge: finding great accounting talent was nearly impossible.
Two years ago, they stopped looking and started building. Debit & Co. launched with two employees and a simple thesis: a small team with the right technology could outperform a larger team running on legacy tools.
Today, they have twelve employees, a growing roster of clients, and a month-end close process that most established firms can't match.
When they launched, Aaron and Kevin evaluated every major platform. They settled on QuickBooks: cloud-based, widely adopted and marketed heavily on its automation capabilities.
"QuickBooks has automations, which is what everyone wants to talk about. But if the automations aren't good, they're actually worse. You end up automating a small problem into a much bigger one."
Aaron watched it happen in real time. A client with employees across 40 states ran payroll, and QuickBooks automatically created new accounts, posted transactions incorrectly, and generated a reconciliation nightmare. Every single month. The automation didn't save time. It created more work.
The firm needed something fundamentally different: software where accountants set the rules and AI executes them exactly, not a black box guessing at intent.
Aaron and Kevin had been early AI adopters, using ChatGPT from day one. They understood the technology and its limits. When they found Puzzle, they recognized what they'd been looking for: an AI-native accounting platform where accountants define the rules and AI executes them.
“We love AI Close. We used it alongside the checklist, working through all the different tasks, and some of what it did was awesome."
The most telling moment came when a team member (not a developer, an accountant) built their own AI Close agent from scratch.
"One person on our team built their own agent within AI Close, and it was very cool. It checks a ton of things for trend analysis. We really enjoyed using it because review is a big part of getting everything out the door."
This wasn't just a feature. It changed what the team could do. Non-engineers could build custom review workflows without IT, without consultants, and without a six-month implementation.
Today, eight members of the Debit & Co. team are Puzzle-certified and running their own closes.
In their first full month with AI Close, Debit & Co. closed every single client by the fourth business day. Not most of them. All of them.
"We were able to shave off between 25% and 30% of the work right off the bat. Across 15 clients, that saves us one to one and a half hours per client."
"We're going to love having 50 clients we can use this on. That's part of the goal for us: to keep growing and adding more clients."
When Aaron describes Puzzle to other firm owners, he doesn't reach for buzzwords. He reaches for the thing that actually matters to a practitioner.
“It's an open template. You can build all kinds of agents and establish different rules. It's like NetSuite, but without paying for everything."
For Aaron, that's the appeal: the flexibility of an enterprise platform without the cost and complexity that usually come with it.
Debit & Co. has already put that flexibility to work, building custom workflows, creating their own agents, and giving more team members ownership of the close process.
They didn't just speed up the close. They built a system that lets a twelve-person team support far more clients without adding headcount at the same pace.
Discover Puzzle.io - the AI-native ledger built to save you time, enhance accuracy, and turn your financial data into actionable insights.
