wtf is oh-my-ai-company?

Oh My AI Company is a continuously updated research database covering AI companies, products, enterprise agents, and vertical AI tools. Rather than a set of scattered reports or a simple ranking list, it works more like a growing market map, connecting companies, products, founders, investors, funding events, traffic signals, community feedback, media coverage, and competitor relationships into a single searchable, traceable, and expandable network of objects. It is built on top of Memex, a local-first knowledge tool created by the same author that stores structured data in SQLite and the actual written content in Markdown files, combining a structured database with readable research notes and a graph of relationships between them. A public website mirrors the whole vault to Cloudflare, keeping all types, objects, fields, and cross references in sync, while stripping out anything sensitive like keys or local file paths before publishing, the local Memex vault always remains the single source of truth. The project deliberately separates facts from opinions: evidence such as official sites, funding reports, and community discussion is stored as source items, personal judgments and takeaways are stored separately as notes, traffic numbers are stored as snapshots, and ongoing monitoring links are stored as touchpoints. As of mid July 2026, the vault holds 87 companies and products, 84 investors, 118 people, and 247 ongoing monitoring touchpoints, spanning areas like enterprise AI agents, browser and automation infrastructure, AI development tools, and vertical AI applications, along with the venture firms and funds backing them. Anyone can browse the Markdown files directly, or use the Memex command line tool or web interface to query the structured relationships between companies, people, investors, and evidence. Running it locally requires building or installing Memex first, then pointing its command line tool at this repository to check status, browse queries, or start a small local web server. The project is explicit that it is a research map meant to be reused and built on, not a finished commercial database or investment advice, and that details like funding, traffic, and team size can go out of date.