Crustdata indexes public web pages and resolves them into one continuously updated graph of people and companies, rather than a static archive of scraped records. That graph is exposed two ways: a flexible REST API with a filter language that supports intent-style queries, and an MCP server that drops the whole graph into an agent in a single line of configuration. Teams use it for search, enrichment, signals and watchers — the four primitives that cover most B2B data workflows.
The product is built around three claims it calls freshness, depth and resolution. Freshness means profiles are re-verified continuously instead of going stale; depth means a profile carries the whole story — role changes, funding events, hiring activity, open-source work — rather than a single row; resolution means every identity is mapped to the right human, which is the hardest part of B2B data and the reason agents so often sound confidently wrong. Crustdata markets itself chiefly to sales, recruiting and investing teams that now run sourcing flows end to end with agents.
Technically, queries are expressed as structured filter conditions over fields such as headcount, headcount growth over twelve months, total funding raised and country, and return records with company name, website, funding and headcount. Because agents do the querying, the interface avoids filter menus and list limits — you describe the target in natural language and the agent translates it into precise filters.
