Full-stack AI web app @ Delta Cubes
Recruitment Management System
Built for Delta Cubes during my 2024–2025 AI/ML engineer role, this AI-powered recruitment platform helps recruiters screen hundreds of resumes in minutes, auto-match candidates to open roles with explainable LLM scoring, and shortlist top-fit talent from a single dashboard — cutting manual screening time by ~70% while keeping decisions transparent.
React.jsNode.jsPostgreSQLOpenAI APILLMs
- Role
- Full-stack + AI engineer @ Delta Cubes
- Timeline
- 2024 — 2025
Problem
Built for Delta Cubes as part of my 2024–2025 AI/ML engineer role, this platform solves a recruiter's core pain: hours spent screening resumes and manually matching candidates to open roles, with inconsistent decisions biased by keyword search. The RMS turns that into a minutes-long, explainable workflow.
Approach
- 01Parsed resumes into structured JSON (skills, experience, education) with an LLM extraction pipeline.
- 02Built a job–candidate matching service that scores fit against role requirements and returns top-k candidates.
- 03Designed a React dashboard for recruiters: search, filters, shortlist, and side-by-side candidate comparison.
- 04Exposed everything through a Node.js + Express REST API backed by PostgreSQL.
Stack
Frontend
- React.js
- TypeScript
- Tailwind
Backend
- Node.js
- Express
- REST APIs
Database
- PostgreSQL
AI / LLM
- OpenAI API
- LLM prompt design
- Resume parsing pipeline
Outcomes
- →~70% reduction in initial resume screening time in internal Delta Cubes testing.
- →Top-k candidate match returned in under ~2s per open role.
- →Structured 100% of incoming resumes into normalized JSON (skills, experience, education) for search & filtering.
- →Explainable match scores let recruiters justify every shortlist decision to hiring managers.
- →Deployed as an internal tool used by the Delta Cubes recruitment team across live req pipelines.