01
Evidence-first
Every important conclusion should trace back to candidate-provided evidence.
Evidence over opaque scores
CogniHire turns candidate claims and interview conversations into grounded evidence that hiring teams can review.
01
Submitted by candidate
02
Extracted statements
03
Questions asked
04
Quoted answers
05
Reviewed by humans
Why CogniHire
01
Every important conclusion should trace back to candidate-provided evidence.
02
AI assists with understanding, questioning and analysis while respecting the information actually provided by the candidate.
03
CogniHire does not replace the hiring decision with an opaque score.
How it works
Step 1
Candidate submits a resume.
Step 2
Claims are read and structured.
Step 3
Questions follow those claims.
Step 4
Reviewers read the trail.
Evidence report
A claim, the question it prompted, the answer it received, and the exact line that supports it.
Candidate claim
“Built REST APIs using Python.”
Source · Resume, line 14
Interview question
Explain how you designed authentication for one of your APIs.
Candidate response
I used FastAPI with short-lived access tokens and refresh tokens kept in httpOnly cookies. Tokens were signed with a rotating secret pulled from our secrets manager, and every route declared an auth dependency, so an expired token never reached a handler. We logged failures per client so we could see brute-force attempts early.
Evidence extracted
“Tokens were signed with a rotating secret pulled from our secrets manager, and every route declared an auth dependency.”
Fictional example · No overall score, no ranking
For companies
Company account
Roles, candidates, interview codes and evidence reports live in one reviewer workspace.
For candidates
You can complete the interview from your browser. We'll check camera and microphone permissions before anything starts, and your answers are what the evidence is built from.