IIT KGP Internship Agent

An AI-powered multi-agent pipeline that automates personalized research internship outreach for IIT Kharagpur students.

Features

• ** CV Parsing**: Extracts skills, projects, interests from PDF/DOCX CVs using Gemini
• ** Faculty Scraping**: Scrapes 188+ faculty profiles from IIT KGP website with real emails
• ** Intelligent Matching**: Auto-selects relevant departments (CS, EC, EE) based on your CV
• ** Research Enrichment**: Uses Serper API to find professor publications & research areas
• ** Personalized Emails**: Generates tailored outreach emails with match scores (0.0-1.0)
• ** Cover Letters**: Creates professional cover letters for each professor

Architecture

┌─────────────────────────────────────────────────────────────────────┐
│                        LangGraph Workflow                            │
├─────────────────────────────────────────────────────────────────────┤
│                                                                       │
│  ┌──────────────┐    ┌──────────────┐    ┌──────────────────────┐   │
│  │   Parse CV   │───►│ Recommend    │───►│   Scrape Faculty     │   │
│  │   (Gemini)  │    │ Departments  │    │   (Playwright)       │   │
│  └──────────────┘    └──────────────┘    └──────────┬───────────┘   │
│                                                      │               │
│                                                      ▼               │
│  ┌──────────────────────────────────────────────────────────────┐   │
│  │                    Enrich Profiles                            │   │
│  │                    (Serper API)                               │   │
│  └──────────────────────────────────────┬───────────────────────┘   │
│                                          │                           │
│                                          ▼                           │
│  ┌──────────────────────────────────────────────────────────────┐   │
│  │               Generate Personalized Emails                    │   │
│  │                       (Gemini)                                │   │
│  └──────────────────────────────────────────────────────────────┘   │
│                                                                       │
└─────────────────────────────────────────────────────────────────────┘

Quick Start

1. Clone & Install

git clone <repo-url>
cd internship-using-ai-agents
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
playwright install chromium

2. Configure API Keys

cp .env.example .env
# Edit .env with your keys:
# - GOOGLE_API_KEY (required for Gemini - get from aistudio.google.com)
# - SERPER_API_KEY (free tier: 2500/month at serper.dev)

3. Run Pipeline

# Basic run with 10 professors
python -m src.main --cv your_cv.pdf --limit 10

# Target specific department
python -m src.main --cv your_cv.pdf --departments "Computer Science" --limit 20

# Set minimum match score
python -m src.main --cv your_cv.pdf --limit 15 --min-score 0.3

Project Structure

devsoc/
├── src/
│   ├── agents/                 # AI Agents
│   │   ├── cv_parser.py        # Agent 1: CV Parsing
│   │   ├── faculty_scraper.py  # Agent 2: Faculty Scraping
│   │   ├── research_enrichment.py  # Agent 3: Research Enrichment
│   │   └── personalization.py  # Agent 4: Email Generation
│   │
│   ├── tools/                  # Agent Tools
│   │   ├── document_parser.py  # PDF/DOCX extraction
│   │   ├── web_scraper.py      # Playwright scraper
│   │   └── search_tool.py      # Serper API integration
│   │
│   ├── graph/                  # LangGraph Workflow
│   │   ├── workflow.py         # Node definitions & edges
│   │   └── state.py            # TypedDict state schema
│   │
│   ├── schemas/                # Pydantic Models
│   │   ├── cv.py               # StudentCV, Project, etc.
│   │   ├── faculty.py          # FacultyProfile
│   │   ├── research.py         # EnrichedProfile
│   │   └── output.py           # EmailOutput
│   │
│   ├── utils/                  # Utilities
│   │   ├── config.py           # Pydantic Settings
│   │   ├── logger.py           # Structured logging
│   │   └── department_recommender.py  # CV → Department mapping
│   │
│   └── main.py                 # CLI entry point
│
├── data/
│   ├── raw/                    # Scraped faculty profiles
│   ├── enriched/               # Enriched with research data
│   └── outputs/                # Generated emails (JSON)
│
├── requirements.txt
├── .env.example
└── ARCHITECTURE.md             # Detailed architecture docs

Configuration

Environment Variable	Description	Required
GOOGLE_API_KEY	Google API key for Gemini	Yes
SERPER_API_KEY	Serper.dev API key (free tier)	Yes
SCRAPER_DELAY	Delay between requests (default: 2.0s)	No
LLM_MODEL	Gemini model (default: gemini-1.5-flash)	No
LOG_LEVEL	Logging level (default: INFO)	No

Sample Output

{
  "professor_name": "Abhijnan Chakraborty",
  "professor_email": "abhijnan@cse.iitkgp.ac.in",
  "department": "CS",
  "match_reasons": ["AI research", "DeepFake detection"],
  "overall_match_score": 0.65,
  "email_subject": "Research Internship Inquiry - AI-Powered Platforms",
  "email_body": "Dear Professor Chakraborty,\n\n..."
}

How It Works

1. CV Parsing

• Extracts text from PDF/DOCX using PyMuPDF/python-docx
• Gemini structures data into: skills, projects, internships, interests
• Identifies relevant departments based on skills (ML → CS, VLSI → EC)

2. Faculty Scraping

• Navigates to department pages (/department/CS, /department/EC)
• Selects "Faculty" from dropdown, waits for dynamic content
• Extracts: name, email, designation, research areas, profile URL

3. Research Enrichment

• Searches Google via Serper API for each professor
• Queries: general info, publications, Google Scholar
• Enhances profiles with recent research, publications

4. Email Generation

• Calculates match score based on CV ↔ Professor overlap
• Gemini generates personalized email referencing:
  • Specific professor research areas
  • Student's relevant projects
  • Common interests/technologies

Performance

Metric	Value
Departments scraped	5 (CS, EC, EE, MA, CH)
Faculty profiles	188 total
Serper searches/prof	3 queries
Email generation	~15s per email
Avg match score	0.55 - 0.70

Development

# Run tests
pytest tests/

# Format code
black src/

# Type checking
mypy src/

🤝 Contributing

We welcome contributions! We have prepared a detailed roadmap of standalone features that are ready to be built. Whether you are a beginner or an expert, there is something for you.

👉 Check out the Community Roadmap & Contribution Guide

Quick Start for Contributors

1. Fork the repository

2. Create a feature branch (git checkout -b feature/amazing-feature)

3. Commit your changes (git commit -m 'Add some amazing feature')

4. Push to the branch (git push origin feature/amazing-feature)

5. Open a Pull Request

License

MIT License - Built for DevSoC IIT Kharagpur

Credits

• LangGraph - Agent orchestration framework
• Playwright - Web scraping
• Serper.dev - Google Search API
• Google Gemini - LLM for NLU/NLG