Welcome to agentdatahub
What we're building, who it's for, and why we shaped it as the data layer for AI agents — not as another B2B sales database.
- → agentdatahub is the data layer for AI agents — six datasets behind one API key.
- → Company data, contact data, contact verification, email finder, email verification, SERP API.
- → REST + MCP-native. Pooled credits across every endpoint.
- → Early access is open — design partners get hands-on onboarding.
The story behind why this exists is short. Every AI agent we tried to build ran into the same wall: the data you need is scattered across five vendors, each with its own API surface, its own credit system, and its own pricing philosophy. Want to find a contact, verify their email, pull their company’s funding history, and run a SERP query? Five contracts. Five credit pools. Five things to keep healthy.
We don’t think that holds in a world where agents — not humans — call tools.
What agentdatahub is
Six datasets, one API key:
- Company data — firmographics, funding, tech stack, signals.
- Contact data — person-level records by title, seniority, department.
- Contact verification — live freshness check on any contact.
- Email finder — forward (name + domain → email) and reverse (email → person).
- Email verification — syntax, MX, SMTP, deliverability, disposable.
- SERP API — live SERP, keyword volumes, backlinks, WHOIS.
Every dataset behind one REST endpoint and one MCP server. Credits pool across the lot.
Why MCP-first
Agent hosts — Claude Desktop, Cursor, Windsurf, n8n — are converging on Model Context Protocol. Drop one config block in, and your agent has a new set of tools. We treat MCP as a primary interface, not a wrapper around a sales API. The MCP server isn’t an afterthought. It’s the shape of the product.
What we won’t build
- A clicky UI for sales reps. Not the buyer.
- Unverified contact data dumped wholesale. Provenance on every record.
- Per-seat lock-in or “contact sales” gates on usage tiers.
- A separate vendor for each dataset.
Where we are now
Early access. We’re onboarding design partners — teams building real agents who can shape the schemas with us before broader access opens. If that’s you, join the waitlist and tell us what you’re building.
The infrastructure under the hood is Clevenio’s — the same data work that already powers production B2B systems in the Nordics and EU. agentdatahub is the agent-native interface to it.
— The agentdatahub team