Inspiration

Job hunting in Pakistan is full of scam listings — fake postings, "recruiters" asking for fees, and offers that never lead anywhere. We wanted a way to flag these before candidates waste their time or money.

What it does

TrustHire analyzes job listings and scores them for authenticity using a hybrid pipeline — rule-based checks combined with an LLM that looks for red flags common in scam postings.

How we built it

• Backend: FastAPI, with a three-layer scoring pipeline
• LLM integration via OpenRouter
• Frontend: React, connected to the backend for real-time scoring

Challenges we ran into

Balancing rule-based heuristics with LLM judgment so the score stays accurate without over-relying on the model (and staying within free-tier API limits).

Accomplishments we're proud of

Getting a working end-to-end pipeline live — from listing input to a real authenticity score — on a free-tier setup.

What's next

Polishing the UI, expanding the rule set, and testing against a wider range of real and scam listings.

Built With

• fastapi
• javascript
• llm
• openrouter
• python
• react
• restapi
• vite