Prospect data your agents can call.
Hubic is a proprietary B2B data layer built for AI. Plug it into ChatGPT, Claude or Gemini over MCP, or call it from your code. Your agent gets verified emails, direct dials and live company intelligence back as clean JSON.
You, in Claude
Find the head of sales at Kestrel Analytics and get me a direct line.
Claude used Hubic enrich_person { "company_domain": "kestrel.example", "title": "head of sales" }
- name
- Maya Okafor
- title
- VP Sales
- maya.okafor@kestrel.examplevalid
- phone
- +1 415 555 0142direct dial
- company
- Kestrel Analytics, 51–200 employees, Series B
- signal
- 6 new sales roles in 30 days
Works with ChatGPT, Claude and Gemini
SDKs for Python, Node, Go and PHP
What your agent can look up
Two kinds of records, people and companies. Every field comes back as structured JSON.
People
- Work email
- With a deliverability status on every address.
- Phone numbers
- Mobile and direct dial, labeled by type.
- Role
- Title, seniority and department.
- Profile
- Location and public profile links.
Companies
- Firmographics
- Industry, headcount range, HQ, founding year, company type.
- Signals
- Funding rounds, leadership changes, launches and expansions, each with a date and a source.
- Hiring
- Open roles by department and how fast they are growing.
- Financials
- Revenue range, total funding, last round and investors.
Three ways to connect
Pick the one that fits where your agent runs. They all return the same data.
MCP server
Add one URL to Claude, ChatGPT or Gemini. Your assistant gets Hubic tools and decides when to use them.
https://mcp.hubic.io/mcpMCP setupREST API
JSON over HTTPS with an API key. Enrich one record or send a batch.
POST /v1/people/enrichAPI basicsSDKs
Typed clients for Python, Node, Go and PHP that handle auth, retries and pagination.
pip install hubicSDK quickstartsBuilt for agents
- Shaped for models
- Responses use stable field names and compact JSON, so an agent can read them without prompt workarounds.
- Status on every contact point
- Each email and phone comes with a status, so the agent knows what is safe to use.
- Dated facts
- Signals and company changes carry the date they happened and a source link, so the agent can tell fresh from stale.
- One source, every runtime
- The same data sits behind the MCP tools, the API and the SDKs.
From API key to a usable contact in a few lines of Python.
from hubic import Hubic
hubic = Hubic(api_key="HUBIC_API_KEY")
person = hubic.people.enrich(
company_domain="kestrel.example",
title="head of sales",
)
print(person.work_email, person.email_status)
# maya.okafor@kestrel.example valid
print(person.direct_dial)
# +14155550142What teams build with it
AI SDR agents
Research an account, find the right person and write the first touch with real context.
CRM enrichment
Fill missing emails, phones and firmographics on new records as they arrive.
Account research
Ask an assistant what changed at a target account this quarter and get dated answers.
Inbound routing
Score and route sign-ups by company size, industry and hiring activity.
Give your agents better data.
Tell us what you are building. We will send an API key and MCP credentials.