Hirewise

• 

  HireWise Sign in — Dark minimal UI with role-based authentication

• 

  Recruiter Dashboard — Post jobs and view AI-ranked candidate list

• 

  Candidate pageCandidate Dashboard — Upload resume and apply to jobs with AI ranking

• 

  AI Rankings Panel — Match scores, Select/Reject/Review decisions with reasoning

Inspiration

Recruiters waste hours manually screening resumes — reading hundreds of PDFs, inconsistently rating candidates, with no clear explanation for decisions. I wanted to build something that fixes this completely.

What it does

HireWise is a full-stack AI recruitment platform that:

• Parses resumes automatically (PDF/DOCX) using LLM
• Ranks candidates against job descriptions with a 0-100 match score
• Delivers explainable decisions — Select, Reject, or Review — with reasoning, strengths, and skill gaps
• Supports 3 user roles — Recruiter, Candidate, Admin

How I built it

• Frontend: React 18 + Vite + Zustand — dark minimal UI with cyan accents
• Backend: FastAPI + Python + SQLAlchemy + JWT Auth
• AI: Groq API with LLaMA 3.3-70B-Versatile for resume parsing and candidate ranking
• Database: SQLite with SQLAlchemy ORM
• Deployed: Netlify (frontend) + Railway (backend)

Challenges

• Getting the LLM to consistently return valid JSON required careful prompt engineering and regex fallback parsing
• Handling Python 3.13 compatibility issues on Railway deployment
• Managing CORS between Netlify frontend and Railway backend
• GitHub secret scanning blocking API keys in commits

What I learned

• Full end-to-end deployment of a React + FastAPI app
• Prompt engineering for structured JSON output from LLMs
• Role-based JWT authentication in FastAPI
• Real-world deployment challenges with Railway and Netlify

What's next

• PostgreSQL + AWS S3 for production scale
• Email notifications for candidates
• Bulk resume upload feature
• Mobile responsive improvements

Built With

• api
• fastapi
• groq
• llama
• netlify
• python
• react
• sqlite