RevorRevor

Prospect discovery

Revor AI continuously finds companies or people that match natural-language criteria and stores them in a prospect list that supports pagination and live updates. The API calls this resource a webset.

Create a new list

  1. Call POST /websets and save webset.id.
  2. Call GET /websets/{id}/items for an authoritative snapshot.
  3. Open Items SSE with the snapshot's stream_cursor + items_revision.
  4. Apply items.changed events until complete.
  5. On reset, replace local state with a new snapshot and reconnect from its checkpoint.

Find more prospects

  1. Read the current Items snapshot and retain the checkpoint from its first page.
  2. Call POST /websets/{id}/find-more and save item.id.
  3. Open Find More SSE with that job_id and the old checkpoint.
  4. Apply changes until complete.

Snapshot first, then stream. This closes the gap between the two requests. A stream opened without a checkpoint starts at the latest event and only delivers future changes.

Endpoints

Method and pathDescription
POST /api/v2/websetsCreate a company or person prospect list.
GET /api/v2/websetsPage through the current user's lists.
GET /api/v2/websets/{id}Get status, criteria, and progress.
GET /api/v2/websets/{id}/strategyRead criteria and search strategy.
GET /api/v2/websets/{id}/itemsPage through an authoritative item snapshot.
GET /api/v2/websets/{id}/items/streamStream state, strategy, and item changes.
POST /api/v2/websets/{id}/find-moreContinue discovery on an existing list.
GET /api/v2/websets/{id}/find-more/streamStream one specific Find More operation.

Permissions

Reads and SSE require websets:read. Create and Find More require websets:write. Resources are isolated by the API key owner's user account; another user's ID returns 404 webset_not_found.

Statuses

Text
draft | generating_criteria | ready | queued | searching | verifying |
completed | failed | cancelled

A preparation_job only reports whether creation or Find More started successfully. The Webset status reports whether prospect discovery is complete.

Two cursor types

  • next_cursor is an opaque REST pagination cursor. Pass it unchanged to the next page.
  • stream_cursor is a numeric SSE event cursor. Store it together with items_revision.
  • The cursor types are not interchangeable.