AI Job Agent
Autonomous AI agent that searches job boards, matches opportunities to your profile, customizes application materials, and tracks your job search pipeline — all without manual effort.
The Problem
Job searching is a full-time job. Manually browsing boards, tailoring CVs, and tracking applications is exhausting and error-prone. Most candidates apply to ill-fitting roles.
The Solution
Built a multi-agent system that continuously monitors job boards (LinkedIn, Indeed, company career pages), scores opportunities against a candidate profile using semantic similarity, generates tailored cover letters, and maintains a searchable CRM for the entire pipeline.
Architecture
FastAPI backend with APScheduler for periodic scraping, LangChain agents for reasoning and generation, pgvector for semantic job matching, and a React dashboard for pipeline visibility.
Key Challenges
Anti-scraping measures on major job boards required building a resilient scraping layer with rotating proxies, rate limiting, and fallback strategies.
Lessons Learned
LLM quality is only as good as your prompt engineering. Structured output with Pydantic validation was essential for reliable agent outputs.
Impact & Results
Reduced job search time from 20+ hours/week to under 2 hours of review. Increased application-to-interview rate by matching only high-fit opportunities.