One plain-English query returns a ranked, structured list of the companies and people that fit, over a REST API or straight into your agent via MCP.
Example queries: seed-stage fintech in NYC hiring engineers. DTC skincare brands doing $5M+ on Shopify. AI infra companies that raised in the last 90 days.
PARSED → MATCHED → SCORED, IN ONE CALL
seed-stage fintech startups in NYC hiring engineers
FIND NEW LEADS · COMPLETE THE ONES YOU HAVE · TRUST EVERY FIELD
A plain-English query returns scored, structured leads. No filters, no exports.
Already have a list? Send an email, get back a complete record, merged across every source.
Every email deliverability-checked, every mobile validated, every field cross-checked. Your list isn't full of bounces.
SOURCED FROM PUBLIC + LICENSED DATA, CROSS-CHECKED ACROSS 2+ SOURCES
HOURS OF WORK → A LIST THAT'S STALE BY MONDAY
GET A KEY → FIRST RESULT BEFORE YOUR COFFEE'S COLD
One POST. JSON in, ranked leads out.
POST /v1/search/companies{ "query": "fintech in NYC" }
Point any agent at it. It just works.
"disburse": { "url": "mcp.disburse.dev"}
TypeScript & Python. Two lines.
import { disburse }await disburse.search("…")
NO SEATS · NO ANNUAL LOCK-IN · START FREE
The index is re-crawled continuously. Most records are under a week old, and funding and hiring signals update daily.
Yes. disburse ships a native MCP server. Point Claude or any agent at it and lead search becomes a tool it can use.
Structured JSON for each match: name, domain, key signals (stage, headcount, tech), and a 0–1 relevance score.
Your queries appear in your own request log so you can debug and see what's running. They're never used for training or shared, and you can turn query logging off entirely in Settings.