Documentation / App / Similar Prospects
Similar Prospects
Apphttps://oppalerts.com/app/opportunities/similar_prospects
What you use it for
- Find the pages nearest in semantic similarity to one page you already have, in the Prospect Index.
- See who republished the same page. A copy on another site is a result like any other, near the top of the list.
- Widen a search from one good result without writing a query.
How it works
The page you came from is looked up by its id, and the stored vector of that document is compared against every document in the Prospect Index. The 100 nearest are returned, nearest first, with one row each in the same format as every other result page.
You can open it from either index. Every result card offers both searches, so this page is opened on a News Index result too. Then the page looks for the same address in the Prospect Index and uses the Prospect Index copy of that page. The search itself always reads the Prospect Index.
Two cards
- This document is the one you opened.
- Nearest documents. The 100 documents nearest to it, nearest first. Copies of the same text on other sites are not merged: one page that ten sites carry takes ten rows.
Similarity is semantic similarity, not a grade. The percentage on a row is the vector distance. Read the order and the gaps between rows, not the number on its own.
On your own campaign
- Start from a result you like on any index page and click "Similar prospects".
- Read the top rows first when the page is a press release: the sites that published the same text are near the top, and that is the list of sites that take that page.
- Click "Similar prospects" again on a nearest document to walk outward one step at a time.
- "The Prospect Index does not hold that document" means the Prospect Index has no copy of that page. It happens when you open this page from a News Index result that the Prospect Index never indexed. Click "Similar news" on that result instead.
MCP
These MCP tools give the same data. Each page has every parameter and an example.
prospects_similar_to_id: The 100 Prospect Index pages nearest in semantic similarity to one page, named by its document_id, nearest first.
API
These API routes give the same data.