Full-stack · BTech mini project
CV Management System.
A recruitment app where candidates apply and recruiters rank them.
Candidates keep one structured profile and apply to jobs. Recruiters see applicants ranked by a score they can adjust. Built by a team of three.
- React
- TypeScript
- Supabase
- PostgreSQL
- TanStack Query
fig. 05
skills ×0.6exp ×0.2role ×0.2skills-first
- 1ACandidate A7.2
- 2CCandidate C6.6
- 3BCandidate B6.2
- 4DCandidate D4.2
What I built
- My contribution: Supabase schema and PostgreSQL database design, TanStack Query integration, and the configurable weighted scoring algorithm.
- Candidate side: CV form, job list, one-click apply and a status tracker.
- Recruiter side: post jobs, review a ranked applicant table, move applicants between stages.
- Weighted score from skill overlap, experience, role match and bonuses, with recruiter-adjustable weights.
- Five related Postgres tables in Supabase with foreign keys and cascading deletes.
- TanStack Query keeps lists and counts in sync after a status change.
Hard parts
CHALLENGE // 01
Ranking without a black box
- Problem
- A hidden score can quietly decide who gets seen.
- Fix
- The score is a plain weighted sum. Everyone stays visible, and the recruiter can see how each factor contributed.
Limits and next steps
- – Matching is keyword-based. Semantic matching (Postgres ≈ SQL) is not built.