The Startup Using AI to Translate Documents Into Data

If you’ve ever uploaded a picture of a receipt to an expense report or read a PDF of a book online, you’ve likely used optical character recognition, a decades-old technique that converts images of typed, handwritten or printed text into text that’s editable on a computer.

OCR might not sound like the sexiest market. But it’s interesting enough for Andreessen Horowitz, one of the most prolific backers of young AI startups in the last two years, to lead a new round of funding for Reducto, a startup that’s combining OCR with newer AI technology to understand documents. The round comes just six months after Reducto announced its last round, led by Benchmark at a $200 million post-investment valuation.

The most recent $75 million Series B round will triple Reducto’s valuation to $600 million, including the investment. This fundraise brings Reducto’s total funding to $108 million, and it still has more than $100 million in the bank, meaning that it’s burned less than $8 million in its two years of existence, cofounder and CEO Adit Abraham told me.

The San Francisco-based startup combines traditional OCR techniques with vision-language models (VLMs), a newer type of AI model that can understand and generate text, images and videos. That helps its tech better understand complex documents in domains like finance, healthcare, legal and insurance and get them into a format that can be fed into large language models, Abraham said.