InternHunt – AI‑Powered Resume & Job Assistant

Parse resumes, extract skills, discover jobs and courses, classify domain, and chat locally with an AI coach — all in a modern Streamlit app.

---

Overview

InternHunt helps candidates quickly understand and improve their resumes, discover relevant roles and courses, and converse with a local AI about improvements. The app is modular and production-ready, with graceful fallbacks when optional services (DB, APIs, Ollama) aren’t configured.

• Live UI with Streamlit
• Resume parsing and skill extraction
• Job and course recommendations
• Resume domain classification
• Local chatbot via Ollama (optional)
• MySQL persistence and admin features (optional)
• Job scraping across multiple sources

---

Features

• Resume Parsing
  • Extracts contact info, skills, education, and experience from PDFs using spaCy and custom rules in resume_parser.py.
• Smart Recommendations
  • Jobs: role and keyword suggestions based on extracted skills.
  • Courses: recommended courses to fill skill gaps (Courses.py, api_services.py).
• Resume Classification
  • Classifies resume domain/type (e.g., Data Science, Development) via resume_classifier.py.
• Job Scraping
  • Aggregates opportunities from multiple sources in job_scrapers.py with deduplication.
  • Scrapers include Internshala, GitHub repos (hiring/internship topics), and RemoteOK.
• Local Chatbot (Optional)
  • chat_service.py integrates with Ollama for fast, local, resume‑aware Q&A.
  • Streaming responses, model health checks, and conversation context.
• Database (Optional)
  • database.py provides MySQL connectivity for admin and tracking features.
  • setup_database.sql bootstraps schema; app runs fine without DB configured.
• Robust UI/UX
  • Clean components in ui.py, theme and styles from styles.py.
  • Centralized error handling via error_handler.py.
  • Config in config.py and helpers in utils.py.

---

Project Structure

.
├─ App_refined.py           # Main Streamlit entrypoint (recommended)
├─ App.py                   # Legacy/alternate entry
├─ api_services.py          # External helpers (yt_dlp, course helpers, Jooble, etc.)
├─ chat_service.py          # Local chatbot via Ollama
├─ job_scrapers.py          # Internshala / GitHub / RemoteOK scrapers + aggregator
├─ resume_parser.py         # Resume parsing & extraction logic
├─ resume_classifier.py     # Resume domain/type classifier
├─ Courses.py               # Course recommendations
├─ config.py                # Env loading & app config (dotenv)
├─ database.py              # MySQL connectivity (optional)
├─ error_handler.py         # Centralized error handling/log helpers
├─ styles.py                # Theming and styled components
├─ ui.py                    # UI building blocks
├─ utils.py                 # Utilities and helpers
├─ setup_database.sql       # DB schema bootstrap
├─ requirements.txt         # Pip dependencies
├─ environment.yml          # Conda environment (alternative to pip)
├─ .env.example             # Template for environment variables (no secrets)
├─ .gitignore
├─ README.md
└─ logo.png

---

Getting Started

Prerequisites

• Python 3.9+
• pip or Conda
• spaCy model en_core_web_sm
• Optional:
  • MySQL Server (for DB features)
  • Ollama (for the local chatbot)

Setup

Option A: pip + venv

python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python -m spacy download en_core_web_sm

Option B: Conda

conda env create -f environment.yml
conda activate internhunt
python -m spacy download en_core_web_sm

Environment Variables

Copy and edit the example:

cp .env.example .env

Minimum variables (aligned with config.py):

# Database (optional)
DB_HOST=localhost
DB_USER=root
DB_PASSWORD=
DB_NAME=cv

# Chatbot (optional)
OLLAMA_HOST=http://localhost:11434
OLLAMA_MODEL=phi:latest

# Job APIs (optional)
JOOBLE_API_KEY=
ADZUNA_APP_ID=
ADZUNA_API_KEY=
ADZUNA_COUNTRY=in

If DB or Ollama are not configured, the app still runs with those features disabled.

Run

Use the Streamlit runner:

streamlit run App_refined.py --server.port 8502

Open the URL printed in the terminal (e.g., http://localhost:8502).

---

Optional Integrations

Database (MySQL)

1. Start MySQL and apply schema:

mysql -u root -p < setup_database.sql

2. Ensure .env contains the correct credentials.

3. Run the app; database‑backed features (admin/tracking) will be enabled.

Local Chatbot (Ollama)

1. Install and run Ollama:

• macOS: brew install ollama
• Start server: ollama serve

2. Pull a model:

ollama pull phi
# or: ollama pull llama3

3. Configure .env:

OLLAMA_HOST=http://localhost:11434
OLLAMA_MODEL=phi:latest

4. Launch the app and open the chat panel.

Job Scrapers

• job_scrapers.py aggregates results from multiple sources and deduplicates by URL.
• Some sources are rate‑limited; the code throttles requests (time.sleep) and handles errors gracefully.
• Provide skills (and optional location) to scrape_all(skills, location).

---

Usage Guide

• Upload a resume (PDF) on the main screen.
• Review extracted contact details, skills, education, and experience.
• Explore recommended job roles and keywords.
• See course recommendations to fill skill gaps.
• Check resume classification (e.g., DS, Dev).
• Use the chatbot for tailored suggestions (if Ollama enabled).
• If DB is enabled, use admin/tracking features as configured.

---

VS Code Debugging

Use a Streamlit launch configuration so Streamlit provides a proper ScriptRunContext:

{
  "name": "Python: Streamlit",
  "type": "python",
  "request": "launch",
  "module": "streamlit",
  "args": ["run", "App_refined.py", "--server.port", "8502"],
  "cwd": "${workspaceFolder}",
  "justMyCode": true,
  "console": "integratedTerminal",
  "env": { "PYTHONPATH": "${workspaceFolder}" }
}

---

Troubleshooting

• Streamlit “missing ScriptRunContext” warnings
  • Run with streamlit run App_refined.py, not python App_refined.py.
• spaCy model not found
  • python -m spacy download en_core_web_sm
• MySQL connection errors
  • Verify server is running; credentials in .env match; database exists via setup_database.sql.
• Ollama connection errors
  • Confirm ollama serve is running; model is pulled; OLLAMA_HOST is correct (no trailing /api, no trailing slash).
• yt‑dlp or API errors
  • Ensure URLs are valid; some endpoints may require API keys configured in .env.

---

FAQ

• Can I run without MySQL?
  • Yes. The app detects missing DB credentials and runs with DB features disabled.
• Which Ollama model should I use?
  • Defaults to phi:latest. You can try llama3 or other models depending on your hardware.
• Do I need API keys?
  • Only for the optional job APIs. Scrapers and the core app work without them.

---

Contributing

Issues and PRs are welcome. For major changes, please open an issue first to discuss your approach.

---

License

MIT. See LICENSE.

---

⭐ If you found this project helpful, please give it a star! ⭐
Built with ❤️ for students and job seekers