Byron

Byron is a San Francisco company, founded in 2025, that builds AI agents for business tax preparation. Founded by Blaze O'Byrne and Wilm Kranz, the team combines artificial intelligence and accounting experience from Amazon's Artificial General Intelligence group and the Big Four firm Deloitte. Byron announced a six and a half million dollar seed round led by Square Peg, with Sorenson Capital, Liquid 2 Ventures, and Correlation Ventures participating, and launched publicly in May 2026 after first customers went live in January. Its thesis is that accounting firms face a capacity problem, not a demand problem, with more than one hundred twenty thousand open roles each year, and that recent advances in large language models finally make it possible to automate the complex business tax process rather than just simple individual returns.

Byron supports 1065, 1120, and 1120-S business returns and works inside the tools firms already use, from email and file storage to accounting systems, Excel, and tax software. A system of AI agents, each trained for a specific task, rolls forward prior year returns and workpapers into a current year Excel workbook, requests missing client information, extracts data from Excel and PDF files, handles partnership allocations and K-1 forms, and produces review ready output. Each extracted number is source linked and confidence scored, and the agents use tools like Python to run the math accurately and reduce hallucinations. The design keeps accountants in control: nothing exports to the tax engine until the firm reviews, approves, and signs off, with source links, confidence scores, approval gates, and audit trails built into every line. Edits made in Excel flow back into Byron without rekeying.

Byron is built for the sensitive data behind every return, with SOC 2 Type II compliance, encryption in transit and at rest, United States based data hosting, role based access controls, and a commitment not to train models on customer data. It integrates with platforms such as Canopy and Truss, which are bringing Byron into their own workflows. It is a strong fit for CPA firms that want to automate the manual preparation behind complex business returns while keeping reviewers firmly in control. It is a weaker fit for firms focused only on individual returns, or those needing open model choice or self hosted deployment, which are not offered.