What it does
A search product over an agency's own candidate database. Past applicants, interview notes and placements are indexed so a recruiter can search the way they think instead of by keywords. Results show why each person matches and how recently they were in touch. The recruiter picks people for a short re-engagement campaign, and replies update each profile.
The problem
Most candidates in the database were never contacted again after their first application. Recruiters paid for new job-board leads while strong past applicants sat unused.
Approach
Resumes and notes are indexed by meaning as well as keywords, so "payments" also finds "UPI" and "card switching". Each result explains its match in one line. Candidates who opted out are never contacted, and the recruiter approves every campaign before it is sent.
- Input
- AI step
- Rules or code
- Person decides
- Result
Screens
Swipe sideways to see the whole screen.
Client work is shown under confidentiality agreements: no client names, and screens are recreated with sample data.
Stack
- Semantic search
- Gemini API
- ATS data sync
- Email and WhatsApp
Impact
Typical for systems like this.
