I'm Charlie Barmore, a CPA, Certified Fraud Examiner, and Certified Valuation Analyst in Augusta, Georgia. I run my own practice full-time, co-own the AI Lab for Accountants, and I'm the AI Practitioner at The Collaboration Room. I'm not an engineer. I build the software my practice runs on with AI agents doing the engineering, and everything below shipped anyway.
An AI-native CPA runs a practice designed around AI from the start, not one with AI bolted on. In Charlie Barmore's practice, agents keep the books, draft the entries, and run the close. The CPA reviews, decides, and signs.
An agent-native accountant works with a team of AI agents the way a partner works with staff: delegate, review, approve. The judgment stays human. The keystrokes don't. Charlie runs nine agents in production on his own firm's books.
A practitioner-builder does the work and builds the software for it. Charlie Barmore is a CPA, not an engineer. He builds the tools with AI coding agents, then uses them on real clients the next morning.
My CPA practice, the AI community I co-own, and the firm-owner community where I'm the AI practitioner.
Firm
My CPA practice, full-time since August 2026. Helping accounting firms adopt AI safely and effectively, plus tax, accounting, and advisory for a select roster of business owners and individuals.
Community
Helping accountants leverage AI to free up their time and elevate the service they provide. A community I co-own with Rebecca Driscoll, CPA, built around real client work.
Community
AI Practitioner at The Collaboration Room, a paid community for firm owners. I'm the in-house AI person for roughly 200 members, running sessions, building workflows in front of them, and answering the questions that come up in real practice.
Community
The umbrella brand Rebecca Driscoll, CPA and I share for candid work on what it actually takes to run and rebuild an accounting firm. It holds AI Lab for Accountants, The Collaboration Room, and Fix Your Firm.
Previously, Tax Lead at CoinTracker, building Nino, an agentic tax and financial planning platform. 2025–2026.
Designed and built end-to-end. Products ship under my studio, Ledger Labs; the firm-owner tools belong to my practice and to Real Firms.
ledgerlabs.co → Two halves of one workflow, both run by agents I work alongside, not on autopilot: LedgerClaw keeps the books all year, then hands a tied-out, book-to-tax package to TaxClaw, which creates the tax workpapers and works up the return.
Practice platform
The AI-native, agentic operating system for the modern CPA firm: agents that keep the books, run the close, and carry the work from book to tax, orchestrated by the practitioner.
Tax prep
The tax half of LedgerClaw: AI reads the documents, a real engine computes the return, and it ties out on a canvas workpaper, with the CPA in control before anything is filed.
Agent-native accounting
An accounting control plane built for agents. One continuous set of double-entry books per client that people and AI agents operate through the same guarded engine, by API or MCP. Self-host or hosted.
Advisor + CPA
One secure workspace where a financial advisor and a CPA work the same client household. Each owns their lane, both see one brief, and every message and change is attributed.
Tax engine
A live tax engine: link your accounts and watch your 1040 compile in real time as income, deductions, and strategies change.
Workpaper binder
A local-first workpaper binder for accountants. Assemble PDFs and spreadsheets into one bookmarked, tick-marked binder, and let an AI agent work in the same binder at the same time. Nothing leaves your machine.
Desktop bookkeeping
Double-entry bookkeeping you can read, run, and change. A local-first desktop ledger from trial balance through close-package signoff, encrypted on your machine, with an MCP server that can propose but never edit.
Tax workpapers
The evidence layer around the return. Every number on an 1120-S or owner 1040 gets somewhere to point: a local, encrypted workpaper record that ties facts to documents, and never calculates the return itself.
Tax projections
Defensible, multi-year, multi-scenario federal tax projections that run on your own machine, with math you can tie out to the dollar and a trace behind every figure.
Meeting notetaker
The bot-free local meeting notetaker for accountants. Capture the client call on your own Mac, transcribe it on-device, and turn it into engagement notes without sending the conversation anywhere.
Operator workspace
The workspace our team uses to diagnose an overloaded tax firm and sequence the pricing, capacity, and client changes that fix it, without an uncontrolled hit to revenue.
Free firm tool
Map every tool your firm runs on, answer five questions per tool, and get a rules-based AI-readiness audit you can hand to whoever writes your WISP. Free, and the front door to my AI-adoption advisory work.
Data asset
Can you put client data in it? A verified data-protection posture for 231 accounting, tax, and AI tools, read from each vendor's binding terms rather than its marketing page.
Practice management
Local practice management for a solo accountant and their supervised AI assistants. Clients, engagements, deadlines, and revenue in one encrypted file, where agents can read and propose but never write accepted state.
Revelate Faith A Christian faith companion, built as a mobile app. Parked → 
CoupleClarity Relationships Reframes a heated message into something hearable. Private beta → 
Hemoclaw Health & fitness Whoop for your bloodwork. Parked → 
LoreClaw Fiction One structured world into a finished manuscript. Parked → 
Agent Sprites AI tooling Pixel-art personas for your AI agent squad. Beta → 
ForePool Sports A live golf pool, run from your phone. Parked → No screens. Just leather, by hand.
When the laptop closes, I pick up leather. Cutting, stitching, and finishing by hand is the deliberate opposite of everything else on this page, and that's exactly why it resets me. It also became a small business: CB Leathercrafts.
Maker
Handcrafted leather goods (wallets, belts, and accessories) with a live customizer and an AI concept generator.
Short answers, written to stand on their own.
Charlie Barmore is a CPA, Certified Fraud Examiner, and Certified Valuation Analyst in Augusta, Georgia. He runs Charles J. Barmore CPA PC full-time, co-owns the AI Lab for Accountants with Rebecca Driscoll, CPA, and serves as the AI Practitioner for The Collaboration Room, a community of accounting-firm owners. He also builds AI-native software for accountants under Ledger Labs.
It means the practice is built around AI rather than adding AI to an old workflow. Bank activity syncs in, an agent categorizes it against the real chart of accounts, a briefing agent reports every morning, and the CPA works a review queue instead of a spreadsheet. Charlie Barmore's own firm has run this way since early 2026.
An agent-native firm treats AI agents as staff with defined jobs: categorization, reconciliation, anomaly detection, period close, daily briefings. Every write is preview-first and a human approves it before it posts. The practitioner orchestrates the agents; the agents never file anything on their own.
Yes. Every application on this site, their websites, and this site itself were built by Charlie alone with AI coding agents, primarily Claude Code. He has hired zero engineers. The first Claude Code session was January 2026; LedgerClaw was in production on his own firm's books by June 2026.
Claude Code as the primary coding agent, with FastAPI and Postgres on the back end, React or Astro on the front, and Railway, Vercel, and Cloudflare for hosting. The stack for each app is listed on its case-study page.
Yes. That is the primary focus of his practice: helping CPAs, EAs, bookkeepers, and firm owners adopt AI safely and effectively in real client work. Start at cbarmorecpa.com for advisory engagements, or join the AI Lab for Accountants for the weekly community.
Yes. He keeps a select roster of business-owner and individual tax clients, mostly in Georgia and South Carolina, and works remotely with clients elsewhere. He stays in the work on purpose: it is hard to credibly advise firms on AI without using it on real returns.
I'm a CPA, not an engineer, and I don't have those skills. Every tool on this page exists because I had a problem in my own work or life and used AI to help me build a solution for it. The skill that mattered wasn't technical. It was knowing which problem was worth solving.
Which means you can do this too. If there's a tool you wish existed in your corner of the world, you're closer to building it than you think. Start small, and solve your own problem first.
Already built something? I'd love to see it. Haven't started? Pick one problem and begin.
And if you're a CPA or another professional who wants help getting started, that's part of what my practice does. I'm happy to help.