Basis, a startup developing artificial intelligence software for accounting firms, today announced that it has closed an early-stage $34 million investment.

The Series A round was led by Khosla Ventures. It also drew contributions from Better Tomorrow Ventures, BoxGroup, Avid Ventures and NFDG, a fund run by former GitHub Chief Executive Nat Friedman and former Apple Inc. executive Daniel Gross. They were joined by several angel investors, including Google LLC Chief Scientist Jeff Dean.

Basis, officially Essex Labs Inc., provides an AI platform that promises to make accountants more productive by automating repetitive parts of their work. The software is powered by language models “built specifically for extremely high accounting performance.” Those models power agents, AI programs optimized to perform a narrow set of actions with little to no human input.

Accountants must often move data between applications as part of their day-to-day work. For example, a finance professional may have to copy the details of a transaction from scanned receipts stored in Dropbox to a centralized ledger. Basis says its AI agents can automate data entry tasks.

Double-checking ledger entries is another chore that the company promises to speed up. Searching for errors manually can be highly time-consuming at a large company that processes upwards of hundreds of transactions per week.

An accounting firm’s clients often manage their financial data in different ways. Two companies that operate in two different markets, for example, may have to structure their earnings reports in different ways to comply with local regulations. Basis says that its AI agents can adapt to each client’s requirements. 

The platform accesses ledger data through integrations with popular accounting applications. According to Basis, its software retrieves records using revocable access tokens and doesn’t store application login details. It leverages several different encryption methods to protect the information it processes. 

Basis says its platform has been adopted by several of the top 100 accounting firms in the U.S. One customer, New Jersey-based Wiss, has reduced the amount of time that staffers spend on manual tasks by up to 30% in some cases.

The company will use the funding to hire more employees. Additionally, the new capital should make it easier for the company to pursue its product development efforts. Both training AI models from scratch and fine-tuning existing algorithms can incur significant infrastructure expenses.

Basis is the latest in a series of startups to have raised funding for AI-powered accounting tools. 

In February, DataSnipper NV received $100 million to boost the productivity of auditors, accountants tasked with checking the accuracy of companies’ financial records. The company’s AI can automatically match a ledger entry to the receipt or other financial document that describes the relevant transaction. More recently, Numeric raised $28 million for a platform that uses AI to help companies understand the cause of fluctuations in their financial performance. 

Image: Unsplash