A personal AI tool that reads job descriptions, matches your best resume, writes tailored emails, and sends them directly from your Gmail — all in one click.
- 📋 JD Parser — Paste text or upload PDF job descriptions
- 🧠 Smart Resume Matching — AI scores and ranks your resumes against the JD
- ✉️ AI Email Writer — Generates professional, tailored job application emails
- 📎 Auto Attachment — Automatically attaches the best matched resume
- 📤 Gmail Integration — Sends directly from your Gmail account
- 📁 Resume Manager — Upload, view and delete resumes from the UI
| Layer | Technology |
|---|---|
| Frontend | React.js |
| Backend | FastAPI (Python) |
| AI / LLM | Groq API (Llama 3.3 70B) |
| Embeddings | HuggingFace Sentence Transformers |
| Vector Store | FAISS |
| Gmail API (OAuth2) |
job-agent/
│
├── backend/
│ ├── main.py # All API routes
│ ├── jd_parser.py # Extract HR email, company, role from JD
│ ├── rag.py # Resume matching with FAISS + Groq scoring
│ ├── email_gen.py # AI email generation using Groq
│ ├── gmail_send.py # Gmail API integration
│ ├── requirements.txt # Python dependencies
│ ├── .env # API keys and personal details
│ ├── credentials.json # Gmail OAuth credentials
│ ├── resumes/ # Your resume PDFs
│ └── vector_store/ # Auto-generated FAISS embeddings
│
└── frontend/
├── src/
│ ├── App.js
│ └── components/
│ ├── JDInput.js # Step 1 - Paste or upload JD
│ ├── ResumeSelector.js # Step 2 - AI resume suggestions
│ ├── EmailPreview.js # Step 3 - Review and send email
│ ├── Success.js # Step 4 - Confirmation
│ └── Resumes.js # Resume management modal
└── package.json
┌─────────────────────────────────────────────────────┐
│ YOUR APP (React) │
│ │
│ Step 1: Paste JD text OR upload JD as PDF │
│ Step 2: AI shows resume match scores │
│ Step 3: Review AI-generated email │
│ Step 4: One click → Email sent with resume! │
└──────────────────────┬──────────────────────────────┘
│
▼
┌─────────────────────────────────────────────────────┐
│ FASTAPI BACKEND │
│ │
│ /parse-jd → Extract HR email, company, role │
│ /match-resumes → Score resumes against JD │
│ /generate-email → Write tailored email with AI │
│ /send-email → Send via Gmail API │
│ /resumes → List all uploaded resumes │
│ /upload-resume → Upload new resume PDF │
│ /resume/{name} → Delete a resume │
└───────┬──────────────────────┬──────────────────────┘
│ │
▼ ▼
┌──────────────┐ ┌─────────────────────┐
│ GROQ API │ │ GMAIL API │
│ │ │ │
│ jd_parser │ │ OAuth2 token saved │
│ rag scoring │ │ Sends from your │
│ email writer │ │ Gmail account │
└──────┬───────┘ └─────────────────────┘
│
▼
┌─────────────────────────────────────────────────────┐
│ RAG PIPELINE │
│ │
│ resume_1.pdf → chunks → FAISS embeddings │
│ resume_2.pdf → chunks → FAISS embeddings │
│ resume_3.pdf → chunks → FAISS embeddings │
│ │
│ JD comes in → Groq AI scores each resume │
│ Returns ranked list with match percentages │
└─────────────────────────────────────────────────────┘
- Python 3.10+
- Node.js 16+
- A Gmail account
- Groq API key (free at console.groq.com)
git clone https://github.com/mmanikandan281/Job-agent.git
cd job-agentcd backend
python -m venv venv
# Windows
venv\Scripts\activate
# Mac/Linux
source venv/bin/activate
pip install fastapi uvicorn python-dotenv langchain langchain-google-genai langchain-community faiss-cpu sentence-transformers pypdf google-auth google-auth-oauthlib google-api-python-client groq python-multipartCreate a .env file inside backend/:
GROQ_API_KEY=your_groq_api_key_here
YOUR_NAME=Your Full Name
GMAIL_ADDRESS=your@gmail.com
YOUR_PHONE=+91 XXXXXXXXXX
GITHUB_URL=https://github.com/yourusername
LINKEDIN_URL=https://linkedin.com/in/yourusername
PORTFOLIO_URL=https://yourportfolio.com- Go to Google Cloud Console
- Create a new project
- Enable Gmail API
- Create OAuth 2.0 credentials (Desktop App)
- Download the JSON file
- Rename it to
credentials.json - Place it inside the
backend/folder
Place your resume PDFs inside backend/resumes/:
backend/resumes/
├── resume_fullstack.pdf
├── resume_python.pdf
├── resume_backend.pdf
└── resume_general.pdf
cd ../frontend
npm install
npm install axiosStart Backend:
cd backend
venv\Scripts\activate
uvicorn main:app --reloadStart Frontend (new terminal):
cd frontend
npm startOr use the one-click startup bat file (Windows):
# Double click start.bat in the root folder- Open
http://localhost:3000in your browser - Paste a job description or upload a JD PDF
- Click "Analyse JD" — AI extracts HR email, company and role
- View resume match scores — AI ranks your resumes
- Select the best resume (or override AI suggestion)
- Click "Generate Email" — AI writes a tailored email
- Review the email and edit if needed
- Click "Send Email" — email sent with resume attached!
| Method | Endpoint | Description |
|---|---|---|
| POST | /parse-jd |
Extract details from JD text |
| POST | /parse-jd-pdf |
Extract details from JD PDF |
| POST | /match-resumes |
Score resumes against JD |
| POST | /generate-email |
Generate tailored email |
| POST | /send-email |
Send email via Gmail |
| GET | /resumes |
List all resumes |
| POST | /upload-resume |
Upload new resume PDF |
| DELETE | /resume/{name} |
Delete a resume |
| Task | Model | Why |
|---|---|---|
| JD Parsing | llama-3.1-8b-instant |
Fast, simple extraction |
| Resume Scoring | llama-3.3-70b-versatile |
Smart understanding needed |
| Email Writing | llama-3.3-70b-versatile |
Best quality output |
- Token Limits — Free Groq tier allows ~100K tokens/day (~30 applications)
- Gmail OAuth — First run opens browser for permission, saved after that
- Resume Storage — Resumes stored locally in
backend/resumes/ - Vector Store — Auto-generated in
backend/vector_store/on first run
Add these to your .gitignore:
.env
credentials.json
token.pkl
venv/
vector_store/
node_modules/- Application tracker (log sent emails)
- Email reply detection
- Job board API integration
- Mobile app (Capacitor)
- Cloud deployment
Manikandan M | MCA Student | Software Developer