How Byron built a zero-accounting-debt finance stack from day one

Byron builds AI agents for accountants. Its founder and CEO, Blaze O'Byrne, spent his career in finance at Credit Suisse and Citibank, picked up a master's in computer science along the way, and did a stint in venture capital. His experience in VC definitely shaped how he runs his own books. "Having spent a lot of my career in venture capital, we always used to get incredibly messy books from companies. When we thought about building our own company, we wanted to make sure that we had those consistent and clean from day one."

The Problem

A company built to automate accounting could not let its own accounting be the exception

Byron set out to be AI-native in every function: sales, marketing, and accounting included. But the reality on the ground did not match the ambition. The company runs two entities, and its books lived in separate systems that had to be reconciled against each other every month. "Our month-end close was super complicated because we were always trying to reconcile books from two different things."

Blaze had seen where that road leads. As an investor, messy financials were a pattern: companies that never invested in their accounting infrastructure early, then paid for it during fundraises, audits, and exits. He treats accounting debt like technical debt. It compounds quietly, and by the time it hurts, it is expensive to unwind.

He also had a filter for the vendor noise. "There's a very big difference between building all of your tooling from the ground up on AI, on true AI platforms, versus taking a legacy system and then just putting AI on top of that."

The Solution

An AI-native ledger, live the next day

Byron evaluated the AI-native accounting vendors and chose DualEntry as the best-in-class fit. The switch was fast. "Next-day migration from my old platform onto this. It's the easiest thing I've ever done in my life."

Onboarding ran through a shared Slack channel with DualEntry's team, who worked each question as it came up, from connecting Byron's existing systems to bringing historical data over.

What Byron runs on now:

• Multi-entity consolidation: both entities in one platform, ending the cross-system reconciliation that complicated every close
• Integrated finance stack: best-in-class tools across every function feeding one ledger, saving significant time at month end
• Natural language reporting: Blaze queries the system conversationally and gets custom reports on expenses and revenue at any point in the month
• Board-ready reporting: DualEntry pulls financials from the income statement and balance sheet and generates charts and graphs that drop directly into board presentations
• AI anomaly detection: expense spikes get flagged for review without clogging approval workflows. "As we move into a world where AI costs are increasing across all of our tools, having anomaly detection that alerts us when those have spiked is incredibly helpful to make sure we can stay on top of and manage our costs."
• MCP server access: Byron is building AI-native workflows across the firm, using tools like Claude to move transactions between systems through DualEntry's MCP server

Results

Byron's close now happens in one platform instead of across two systems. Investor and customer compliance questions get answered from clean, current books rather than a reconstruction project. And the finance function is built to scale the way the rest of the company does. "One of the things that excites us about DualEntry is we believe that we can build a really large business with a very small finance team using a product like this."

His advice to other founders is the lesson he took from the other side of the table: "Putting in a best-in-class finance system early makes sure that you don't end up with accounting debt."

"DualEntry represents the future of what AI can bring to the accounting profession."