Back to projects
Case study · Backend
JobSense AI
Autonomous job-hunting agent that continuously searches job boards, scores listings against resumes using Groq AI, and delivers the strongest matches directly to WhatsApp.
Problem
Job searching across multiple portals is repetitive and time-consuming — candidates miss relevant roles buried in noisy listings.
Solution
Automated a daily pipeline that scrapes job boards, scores each listing against the user's resume with Groq AI, and sends a ranked digest via WhatsApp.
Architecture
- —Node.js + Express REST API with QStash background jobs
- —Redis-backed job queue and caching for deduplication
- —PostgreSQL for jobs, users, and match history
- —Groq API (llama-3.1) for semantic resume-to-job fit scoring
- —Arcjet for API protection and rate limiting
Challenges
- —Handling rate limits across scrapers and AI API calls
- —Designing idempotent daily runs without duplicate notifications
- —Solving production issues: malformed AI responses, credential rotation, WhatsApp delivery limits
Outcomes
- —Fully automated daily job intelligence pipeline
- —WhatsApp delivery with ranked job matches
- —Production-grade backend with caching and structured data
Stack
Node.jsTypeScriptRedisPostgreSQLGroq AIDocker