Bring AI-powered candidate search to your users

Using the Endorsed Platform, you can bring next-generation AI search to your users. This enables them to use advanced, natural language queries like, "experience working in consumer applications, west coast timezone". This provides a much better experience over existing boolean search filters. This document covers string-based search. See our other guides for searching via other methods.

There are two main candidate search use cases you can build with the Endorsed Platform:

1. Search all candidates who have applied to a given job

2. Search all candidates, including past candidates

Once set up, searching candidates in your database is as easy as using our GET /candidate_rankings endpoint. See Candidate Querying. To control whether you're searching for new (for a specific job) or existing (all past candidates) new candidates, simply provide a non-null or null value for filter_job_id.

Here are the steps to build candidate search in your application:

1. Obtain an API key for your organization. When you sign for our Platform plan, we'll provide this to you.

2. One-time batch data import. Import relevant Jobs, Nomination Stages, Candidates, Nominations, and Attachments into the Endorsed Platform. Follow this guide at Data Import. Now already-applied candidates will show up in Endorsed rankings and summaries.

3. Verify data import. Check that all data looks good in the Endorsed Platform by visiting the Endorsed Dashboard for your organization. You'll see the Candidates ranked per job, and can issue search queries against them. You can also verify by issuing search queries against our Rankings API, perhaps with curl or Postman.

4. ATS import connection. Once your initial data set has been imported, build a runtime import step into your application. This is so new candidates will show up in the rankings and summaries from Endorsed. We recommend when a new candidate applies, fire off an import request to Endorsed. As new candidates apply, you can see them in the Endorsed Dashboard.

5. ATS query connection. Integrate ranked Candidate search into your application. Route the user's search query into GET /candidate_rankings?q=[search query], and then display the results. Follow the guide here: Candidate Querying.

6. Decorate data. GET /candidate_rankings is designed to return as quickly as possible, and as such it responds only with entity IDs. To decorate these entity IDs and display the full Candidates in your UI, you can either inflate the data with your database or use our REST API to decorate the Candidate Applicants / Summaries.

   1. Coming soon: We'll provide a GraphQL API to return a fully-decorated and ranked search result in one query. Let us know if you're interested in this.

7. Summarize candidates (optional). Call our Summaries API to show advanced AI-powered summaries to your users, to save them time when reviewing candidates.