Instabase

Instabase was founded in 2015 by Anant Bhardwaj, who left a Computer Science PhD program at MIT (after earlier completing an MS in Computer Science at Stanford) to move to San Francisco and build the company full-time. Bhardwaj had spent two summers as a research intern on Google's Search team while at MIT, and drew on his database-systems research background (his MIT advisors included well-known database researchers) to build what was initially pitched as a 'social network for data' -- a platform for uploading, querying, and sharing datasets. Greylock Partners and New Enterprise Associates backed the company at seed stage in August 2015 with $3.75 million, making Greylock one of Instabase's earliest and longest-tenured institutional investors.

Over the following years the product pivoted from a general data-sharing platform toward enterprise automation of unstructured, document-based workflows -- a large, underserved problem in regulated industries like banking, insurance, and healthcare, where mortgage applications, KYC files, invoices, and claims arrive as scanned PDFs and images rather than structured data. Instabase raised a $23.2 million Series A in June 2017 led by Andreessen Horowitz (partner Martin Casado), and acquired Cloudstitch, a spreadsheet-backed web app platform, in February 2018 to extend its low-code application-building capabilities. By 2019-2021 the platform had moved to a transformer-based deep-learning architecture with pre-trained foundation models purpose-built for document understanding, and the company reported that four of the five largest U.S. banks were using its platform.

A $105 million Series B in October 2019, led by Index Ventures (with Spark Capital, Tribe Capital, SC Ventures, and Glynn Capital participating), pushed Instabase's valuation above $1 billion, making it a unicorn roughly four years after founding. In June 2023, Instabase raised a $45 million round (widely reported as its Series C) led by Tribe Capital with participation from a16z, NEA, and Spark Capital, at a valuation that had doubled to roughly $2 billion. That round coincided with the launch of 'AI Hub,' a marketplace of pre-built AI applications incorporating large language models and generative AI, including an integration with OpenAI's GPT-3.5/GPT-4 models -- reflecting the broader industry shift toward LLM-native document intelligence and, later, 'agentic automation' positioning.

Business model

B2B SaaS / enterprise software; platform + application/agent marketplace (AI Hub) sold primarily to large regulated enterprises (banks, insurers, government).

Enterprise software subscription/licensing (platform + usage), likely supplemented by professional services for implementation; exact pricing not publicly disclosed.