Basis Raises $100M for AI Agents at $1.15B Valuation [US]
AI accounting startup Basis announced a $100 million Series B funding round on 24 February, reaching a $1.15 billion valuation. Accel led the round, with Google Ventures, Lloyd Blankfein, and Khosla Ventures participating.
The New York-based company has now raised $138 million total. Founded by Matthew Harpe and Mitchell Troyanovsky, Basis builds autonomous agents that complete accounting workflows end-to-end, from tax returns to audit processes.
Market Penetration
Basis reports deployment across approximately 30% of the top-25 US accounting firms. Client firms claim 20-50% efficiency gains across practices. The company recently demonstrated the first AI agent to autonomously complete an end-to-end 1065 tax return—a complex task requiring coordination across multiple accounting processes.
Unlike chatbots layered atop existing systems, Basis constructed what it describes as unified accounting intelligence. Agents operate continuously in the background, coordinating tasks and returning completed deliverables for accountant review.
What the Money's For
Basis will deploy capital toward platform development and expanding engineering and machine learning teams. The company positions accounting as infrastructure facing growing complexity from regulatory requirements and autonomous economic agents that will eventually require financial reporting.
"Basis is years ahead in accounting AI, and we believe it has what it takes to define this category as it matures," said Miles Clements, partner at Accel.
Vinod Khosla of Khosla Ventures stated the expectation that "in 2026, Basis will drive the same step-change in accounting that software engineering saw in 2025."
Practitioner Implications
For accounting firms, this signals continued pressure to adopt AI-driven automation or risk efficiency disadvantages. The 30% adoption rate among top-25 firms suggests early-mover positioning. The reported 20-50% efficiency gains represent material cost structure implications for practice economics.
The autonomous 1065 completion extends beyond routine data entry to complex, judgement-requiring deliverables. That's genuinely notable if you care about whether machines can actually do substantive accounting work (we do).