Kadoa (kadoa.com) builds the web data layer for finance: AI coding agents that build, monitor and repair deterministic ETL pipelines, turning websites, PDFs and other unstructured public sources into reliable datasets for investment firms. Instead of engineers hand-writing a brittle scraper per source, an analyst points Kadoa at a source and the platform extracts, transforms and loads the data into a spreadsheet or warehouse, with self-healing scrapers and audit trails underneath. The company says it serves many of the world's top investment firms, and a June 2026 company post describes the client base as tier 1 multi-strategy funds, market makers, private equity firms, investment banks and sovereign wealth funds. Compliance is built in: automated controls for robots.txt, PII bans and source allow lists, with full audit logs of who scraped what and when.
Co-founders Adrian Krebs and Johannes Engler are engineers who worked together at a Swiss bank and experimented with GPT-3 for generating and maintaining web scraper code; their product search engine side project gained traction on Reddit and Hacker News, and Tavis Lochhead (ex Cisco and eBay) reached out and joined as the third co-founder. Kadoa launched publicly with a Show HN for fully autonomous GPT-4 web scrapers in April 2023 and reported reaching CHF 100K ARR within 8 months. The company is headquartered in Zurich, Switzerland with a distributed team; its LinkedIn profile counts 13 employees as of 2026.