An intelligent, autonomous job recommendation agent that parses your resume and searches LinkedIn, Indeed, and Google Jobs for the perfect matching open roles — powered by LangGraph, JobSpy, and your choice of Google Gemini or OpenCode Zen (free models).

🎥 Video Demo

https://github.com/Aryan-Pardeshi/Job-Recommendation-LangChain/raw/main/images%20and%20vids/2026-03-19%2013-47-19.mp4

---

✨ Features

Core

• Resume Parsing Engine — Upload your PDF resume directly into the app.
• Agentic Analysis — Uses LangGraph to analyze your career history with your choice of Google Gemini or OpenCode Zen (free models like Big Pickle, MiMo V2 Pro, MiniMax M2.5).
• Multi-Platform Search — Searches LinkedIn, Indeed, and Google Jobs simultaneously via JobSpy (free, no API key).
• Smart Location Detection — Automatically detects your IP and pulls nearby jobs via the free ip-api.com service.
• Manual Preference Overrides — Override AI detections with dropdowns for Work Type, Experience Level, Location, and Job Boards.
• Broader Second Pass — The AI automatically performs a second broader search to find more opportunities.
• Application Tips — Dynamically generates actionable interview and application tips tailored to your resume.

UI/UX

• Job Result Cards — Clean card-based layout with company, location, salary, and source badges.
• Source Badges — Color-coded badges for LinkedIn (blue), Indeed, and Google.
• Salary Display — Formatted salary ranges with currency and interval.
• Expandable Descriptions — View full job descriptions inline.
• Stats Dashboard — See total jobs, remote count, salary ranges, and top companies at a glance.
• CSV Export — Download all results as a CSV file.
• Raw Results Tab — Browse unfiltered results alongside AI recommendations.
• Session State — Preserves results across UI interactions.
• Sidebar Info Panel — Quick access to how it works and what's powering the app.

Architecture

• MCP Server Ecosystem — Decoupled Model Context Protocol server (mcp_server.py) exposes job search tools. The LangChain agent connects via MultiServerMCPClient.
• Three MCP Tools — search_jobs_tool (filtered), search_jobs_broad_tool (unfiltered), list_supported_sites.
• Zero Cloud Dependencies — Everything runs locally. No paid APIs for scraping.
• Dual Provider Support — Choose between Google Gemini or OpenCode Zen (free models) in the UI. API keys are saved to .env automatically.

---

🚀 How to Setup

1. Clone the Repository

git clone <your-repository-url>
cd Job-Recommendation-LangChain

2. Install Dependencies

This project uses uv for fast package and virtual environment management.

# First, install uv globally if you haven't already
pip install uv

# Sync and install the project environment
uv sync

3. Install JobSpy

JobSpy's numpy pin conflicts with Python 3.13, so install it manually:

pip install python-jobspy --no-deps
pip install beautifulsoup4 markdownify regex tls-client

4. API Setup & Configuration

You need an API key for your chosen LLM provider.

1. Copy the example .env file:

   cp .env.example .env

2. Choose your provider:

   Option A — Google Gemini

   • Go to Google AI Studio and generate an API key.
   • Set GOOGLE_API_KEY in your .env file.

   Option B — OpenCode Zen (Free)

   • Go to OpenCode Zen Auth and get your API key.
   • Set OPENCODE_ZEN_API_KEY in your .env file.
   • Available free models: big-pickle, mimo-v2-pro-free, mimo-v2-omni-free, minimax-m2.5-free, nemotron-3-super-free.

   Tip: You can also enter your API key directly in the app UI — it will be saved to .env automatically.

5. Run the Application

uv run streamlit run main.py

Open your browser to http://localhost:8501, upload your resume, and click Find Matching Jobs!

---

🏗️ Project Structure

├── main.py                 # Streamlit UI (entry point)
├── mcp_server.py           # FastMCP server with 3 job-search tools
├── pyproject.toml          # Project metadata and dependencies
├── .env                    # Environment variables (GOOGLE_API_KEY, OPENCODE_ZEN_API_KEY)
├── README.md               # This file
│
├── src/
│   ├── agent.py            # LangGraph agent (LLM orchestration)
│   ├── job_api.py          # JobSpy wrapper (search, stats, mappings)
│   └── fetch_location.py   # IP geolocation via ip-api.com
│
└── images and vids/
    ├── logo.png            # App logo
    └── 2026-03-19 13-47-19.mp4  # Demo video

🔧 Tech Stack

Component	Technology
Frontend	Streamlit
AI Agent	LangGraph + Google Gemini or OpenCode Zen (free models)
Job Scraper	JobSpy (LinkedIn, Indeed, Google Jobs)
Tool Protocol	MCP (Model Context Protocol) via FastMCP
Resume Parser	PyPDFLoader (LangChain)
Location	ip-api.com (free)
Package Manager	uv

📊 JobSpy Data Schema

Each job result includes:

• title, company, site (source platform)
• job_url, location, description
• is_remote, job_type (fulltime, parttime, contract, internship)
• min_amount, max_amount, currency, interval (salary)
• date_posted, company_url, emails