JobHawk

The AI-Powered Job Application Agent

A semi-autonomous AI agent that discovers jobs, scores them against your resume, drafts tailored cover letters, and lets you review & approve applications—all from your pocket.

---

Features

• AI Resume Parsing: Automatically extracts skills, experience, and education using Llama/Gemini.
• Tinder-Style Job Swiper: Review AI-generated job drafts on the mobile app. Swipe left to pass, swipe right to approve.
• Smart Job Matching: Uses vector embeddings and LLM analysis to find jobs that perfectly align with your profile.
• Application Tracking: A dedicated Kanban board to track your application pipeline (Saved, Applied, Interview, Offer).

Architecture

graph LR
    A[Mobile App - Expo] -->|REST API| B(FastAPI Backend)
    B -->|PostgreSQL + Auth| C{Supabase}
    B -->|LLM Prompts| D[Groq / Gemini]
    B -->|Scraping| E[JobSpy / APIs]

Loading

Tech Stack

• Mobile: React Native, Expo, Reanimated, Expo Router
• Backend: Python, FastAPI, Uvicorn, LangGraph
• Database: Supabase (Postgres, Row Level Security, Auth, Storage)
• AI/LLMs: Groq (Llama 3.3 70B), Google AI Studio (Gemini 2.0 Flash)

---

Quick Setup

1. Supabase Database

1. Create a project at supabase.com.
2. Go to SQL Editor and run supabase/migrations/001_initial_schema.sql.
3. Go to Authentication → Providers and enable Email Auth.
4. Go to Storage and create two private buckets: resumes and drafts.

2. Backend API

The intelligent engine powered by FastAPI and AI models.

cd backend
cp .env.example .env
# Add your SUPABASE_URL, SUPABASE_ANON_KEY, and API keys to .env
pip install -e .
python scripts/seed_companies.py
uvicorn main:app --host 0.0.0.0 --port 8000 --reload

3. Mobile App

The React Native app for swiping on jobs and tracking progress.

cd mobile
cp .env.example .env
# Set EXPO_PUBLIC_BACKEND_URL to your computer's local IP (e.g., http://192.168.1.10:8000)
# Set EXPO_PUBLIC_SUPABASE_URL and EXPO_PUBLIC_SUPABASE_ANON_KEY
npm install
npx expo start -c

---

Roadmap

• Phase 1: Foundation — Resume upload, DB schema, native mobile app shell.
• Phase 2: Job Discovery — Multi-source job scraping, ATS parsing, and LLM scoring pipeline.
• Phase 3: Mobile Swiper — Tinder-style UI for reviewing and approving AI drafts.
• Phase 4: Automated Follow-ups — Email integration, calendar tracking, interview prep.
• Phase 5: Analytics — Funnel charts, interview-to-offer conversion metrics.