Basis Reaches Unicorn Status with $100M Series B: How AI Agents Are Reshaping Accounting

On February 24, 2026, Basis — an AI agent platform built specifically for accountants — announced a $100 million Series B funding round at a $1.15 billion valuation, joining the exclusive ranks of AI unicorns less than three years after being founded. The round was led by Accel, with participation from GV (Google Ventures), former Goldman Sachs CEO Lloyd Blankfein, and existing investor Khosla Ventures. Miles Clements from Accel is joining the Basis board.

The accounting industry is not typically the first sector that comes to mind when people think about AI disruption. But Basis is making a compelling argument that it should be — and the caliber of investors backing this round suggests that some of the most sophisticated technology and finance investors in the world agree. Here is a complete picture of what Basis is, what it is doing, why this round matters, and what it signals for the accounting profession and for AI's role in finance more broadly.

What Basis Actually Does

Basis was founded in 2023 by Matthew Harpe and Mitchell Troyanovsky with a focused thesis: accounting is one of the most process-intensive, documentation-heavy, and deadline-driven professions in existence — and the vast majority of what accountants spend their time doing is exactly the kind of structured, rule-governed, document-processing work that AI agents are uniquely well-suited to handle.

The Basis platform deploys AI agents that are specifically designed to understand accounting — not general-purpose AI models applied to financial tasks, but agents trained on the rules, workflows, and client-service patterns of accounting practices. Those agents can:

• Execute client accounting services workflows end-to-end — processing transactions, reconciling accounts, generating reports — without step-by-step human direction
• Handle tax preparation workflows — including, most recently, generating complete 1065 partnership tax returns autonomously end-to-end, the first AI system to accomplish this
• Support audit practices — gathering documentation, testing controls, flagging anomalies, and drafting workpapers
• Learn client-specific patterns — adapting to the particular accounting methods, chart of accounts structures, and preferences of individual clients over time

The platform is already being used by approximately 30% of the top 25 accounting firms — a market penetration figure that, for a three-year-old startup, signals genuine product-market fit at the enterprise level rather than early-stage experimentation.

The 1065 Tax Return Milestone: Why It Matters

The headline technical achievement in Basis's recent announcements is the demonstration of the first AI agent capable of autonomously completing an end-to-end 1065 tax return — the partnership income tax return required for LLCs, partnerships, and other pass-through entities.

A 1065 return is not a simple form. It requires synthesizing information from multiple sources — partner capital accounts, K-1 schedules, income and expense classifications, depreciation schedules, state apportionment factors — and applying a complex set of tax rules that vary by entity structure, industry, and jurisdiction. For a human tax preparer, a 1065 return for a mid-complexity partnership might take 10-40 hours depending on the entity's size and documentation quality.

An AI agent completing this end-to-end — not assisting a human preparer, but executing the return autonomously — is a meaningful technical and commercial threshold. It means the scope of what Basis agents can handle is no longer limited to data gathering, reconciliation, and documentation support. It extends to the full preparation of one of the most complex individual tax products in practice.

For accounting firms evaluating AI adoption, this matters in concrete terms. Partnership returns represent a significant portion of tax practice revenue and an even more significant portion of staff time at peak season. A system that can handle them end-to-end changes the economic model of a tax practice — not by replacing accountants, but by dramatically changing the ratio of returns a firm can process per staff member.

The Investors: Why This Round Is Significant Beyond the Dollar Figure

The investors in Basis's Series B are worth examining individually, because each brings a specific form of validation that goes beyond financial backing.

Accel is one of the handful of venture capital firms with a multi-decade track record of leading rounds in category-defining B2B software companies — the firm backed Slack, Atlassian, Dropbox, and dozens of other enterprise software leaders. When Accel leads a Series B and puts a partner on the board, it is making a thesis bet that Basis can be a generational enterprise software company, not just a well-funded startup.

GV (Google Ventures) brings both financial backing and strategic proximity to Google's AI infrastructure and tools — relevant for a company building agents that need to handle large volumes of document processing and financial data.

Lloyd Blankfein, the former CEO of Goldman Sachs, is one of the most credentialed finance executives alive. His personal investment in Basis is a form of domain validation that is difficult to replicate through institutional backing alone: a person who ran one of the world's preeminent financial institutions believes that agentic AI will transform professional finance and accounting services at scale.

Khosla Ventures (existing investor) has been an early backer of some of the most ambitious AI infrastructure companies in the current cycle. Their continued participation in the Series B signals sustained conviction rather than early-stage hope.

Taken together, this investor group reflects a consensus that AI agents in professional accounting services are not a niche market — they are a large, defensible, and rapidly expanding category.

The Problem Basis Is Solving: Talent Shortage and Rising Complexity

Understanding why Basis is attracting this level of investment requires understanding the structural pressures the accounting industry is facing independently of AI.

The accounting profession is experiencing a sustained talent shortage that has been building for over a decade. The number of students sitting for the CPA exam has been declining year-over-year, experienced accountants are retiring faster than new professionals are entering the field, and the workload per staff accountant is rising as clients grow more complex, regulatory requirements expand, and audit standards become more demanding.

The result: accounting firms are being asked to serve more clients with more complex needs with fewer staff — a structural bind that AI agents are in a uniquely strong position to address, because the work is highly structured, rule-governed, and heavily documented.

Basis is not positioning itself as a replacement for human accountants. It is positioning its agents as the leverage mechanism that allows existing accounting professionals to handle dramatically higher volumes of work — shifting their time from data entry, reconciliation, and routine document preparation to the judgment-intensive, relationship-intensive work that clients actually need human professionals for.

For accounting firms, the economic model is compelling: if an AI agent can handle the first 70-80% of a tax return or audit workpaper, the senior accountant's job becomes reviewing, adjusting, and approving — work that is more valuable, more interesting, and more sustainable as a career than spending tax season manually entering K-1 data.

What Basis Signals for AI in Professional Services

Basis reaching unicorn status in accounting is not an isolated event. It is part of a broader pattern in which AI agents are moving from horizontal general-purpose tools to vertical domain-specific platforms — systems that understand the specific rules, workflows, and professional standards of a particular industry deeply enough to handle complex, high-stakes work with genuine reliability.

This vertical AI trend is playing out across professional services: - Legal: AI agents that draft contracts, conduct due diligence, and manage case documentation - Healthcare: AI systems that handle clinical documentation, prior authorizations, and coding - Finance: AI agents for compliance monitoring, portfolio management, and financial reporting - Accounting: Basis and competitors building agents that understand GAAP, tax code, and audit standards

The companies winning in each vertical are those that have built deep enough domain expertise into their agents that professionals in those fields trust the output — not just as a starting point, but as a reliable foundation for their own judgment and review. Basis's penetration of 30% of the top 25 accounting firms before its Series B suggests it has cleared that trust threshold in a demanding professional environment.

For professionals in any knowledge-intensive field, Basis's trajectory illustrates the pattern worth watching: AI agents are not arriving as general tools that need to be applied carefully to professional tasks. They are arriving as domain specialists — trained, refined, and trusted in specific professional contexts — that change what a professional's day actually looks like.

If you want to understand how to work alongside, direct, and build with domain-specific AI agents in your own professional context, FireStart's Applied AI & Automation Program is built for exactly that. Explore our Guides library with Ember AI to get started, or enroll in Cohort 3 for structured instruction, live sessions, and professional certification.