Multi-Agent Job Search System with CrewAI and Python

An intelligent, AI-powered application that automates the entire job application process for US federal positions (USAJobs). This tool uses CrewAI agents and Google Gemini to analyze job descriptions, tailor your resume, generate personalized cover letters, and compose professional outreach messages—all in one streamlined interface.

🎯 Features

• Job Search: Search USAJobs by keyword and location
• JD Analysis: AI agents automatically analyze job descriptions and extract key requirements
• Resume Tailoring: Customize your resume summary to match specific job requirements
• Cover Letter Generation: Create personalized, professional cover letters for each application
• Outreach Messaging: Generate customized initial contact messages for networking and applications
• Application Tracking: Maintain a log of all applications and outputs
• User-Friendly UI: Streamlit-based web interface for easy job hunting workflow

🏗️ Architecture

The application uses a multi-agent architecture powered by CrewAI:

┌──────────────────────────────────────┐
│     Streamlit Web Interface (app.py) │
└────────────────┬─────────────────────┘
                 │
         ┌───────▼────────┐
         │  Orchestrator  │
         └───────┬────────┘
                 │
    ┌────────────┼────────────┐
    │            │            │
┌───▼───┐   ┌───▼────┐  ┌───▼─────────┐
│  JD   │   │ Resume │  │ Messaging   │
│Analyst│   │   &    │  │   Agent     │
│Agent  │   │ Cover  │  │             │
│       │   │Letter  │  │             │
└───────┘   │Agent   │  └─────────────┘
            └────────┘

Tech Stack

• Backend: Python 3.12
• AI Framework: CrewAI + LangChain
• LLM: Google Gemini 2.5 Flash (via LiteLLM)
• API Integration: USAJobs API
• Web UI: Streamlit
• Data Management: Pandas, CSV logging

Prerequisites

• Python 3.12+
• Google Gemini API key (see instructions below)
• USAJobs API key (optional, but recommended — see instructions below)

How to get API keys

Below are quick links and short step-by-step instructions for obtaining the two API keys used by this project. The links point to a helpful Educative walkthrough that covers both getting started with Google AI Studio and the USAJobs API.

• Google Gemini (AI Studio) — Educative guide & AI Studio

  • Link: https://www.educative.io/projects/build-a-job-hunt-agent-workflow-with-crewai-and-python/g7M1PmZM2L6#:~:text=over%20to%20the-,AI%20Studio,-and%20log%20in
  • Steps:
    1. Open the Educative guide and follow the "AI Studio" section that points you to Google AI Studio.
    2. Sign in with your Google account and open AI Studio (https://studio.ai.google/ or https://aistudio.google.com).
    3. In AI Studio, go to the API keys or credentials area (sometimes under Project settings or Account settings).
    4. Create an API key (or a service account key if required by your org). Copy the key.
    5. Add the key to your project root .env as: GEMINI_API_KEY=YOUR_KEY_HERE.

• USAJobs API — Educative guide & USAJobs API instructions

  • Link: https://www.educative.io/projects/build-a-job-hunt-agent-workflow-with-crewai-and-python/g7M1PmZM2L6#:~:text=move%20to%20the-,API%20request%20page,-.
  • Steps:
    1. From the Educative guide, follow the API section that links to USAJobs developer/API pages.
    2. Visit the USAJobs developer portal and register an application (you'll need a USAJobs account).
    3. After registering, the portal will provide an API key to use for requests.
    4. Add the key to your project root .env as: USAJOBS_API_KEY=YOUR_KEY_HERE.

If you prefer direct vendor docs, use these official starting points:

• Google AI Studio: https://aistudio.google.com/app/apikey
• USAJobs API: https://developer.usajobs.gov/

Quick Start

1. Clone & Setup

git clone https://github.com/Lakshmiec/job_hunt_assistant.git
cd job_hunt_assistant
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

2. Install Dependencies

pip install -r requirements.txt

3. Configure Environment Variables

Create a .env file in the project root:

GEMINI_API_KEY=your_google_gemini_api_key_here
USAJOBS_API_KEY=your_usajobs_api_key_here
EMAIL_ADDRESS=your_email@example.com

4. Run the Application

Option A: Web Interface (Recommended)

streamlit run app.py

Then open http://localhost:8501 in your browser.

Option B: Command Line

python orchestrator.py

📖 Usage Guide

Via Web Interface

1. Enter Search Criteria

   • Job keyword (e.g., "business analyst")
   • Location (e.g., "New York")

2. Paste Your Resume

   • Include your full resume text

3. Add Your Bio

   • Brief professional summary for outreach tone

4. Search & Select

   • Click "Search Jobs" to fetch matching positions
   • Check the jobs you want to apply for

5. Apply

   • Click "Apply to Selected Jobs"
   • Watch as AI agents analyze, tailor, and generate documents
   • Download generated cover letters and track applications

Output Files

All generated materials are saved to data/:

• report.md - JD analysis report
• resume_agent_output.txt - Tailored resume and cover letter
• cover_letters/ - Individual cover letter files by position
• applications_log.csv - Application tracking log

📁 Project Structure

job_hunt_assistant/
├── app.py                      # Streamlit web interface
├── orchestrator.py             # Main pipeline orchestrator
├── usajobs_api.py             # USAJobs API wrapper
├── requirements.txt           # Python dependencies
├── .env                       # Environment variables (gitignored)
├── agent/
│   ├── jd_analyst.py          # Job description analysis agent
│   ├── resume_cl_agent.py     # Resume & cover letter agent
│   └── messaging_agent.py     # Outreach messaging agent
├── utils/
│   ├── config.py              # Configuration & env loading
│   ├── tracking.py            # Application logging
│   └── .env                   # Local .env copy (gitignored)
└── data/
    ├── sample_resume.txt      # Example resume
    ├── applications_log.csv   # Application history
    ├── report.md              # Analysis reports
    ├── resume_agent_output.txt
    └── cover_letters/         # Generated cover letters

🤖 AI Agents Overview

1. JD Analyst Agent

• Role: Job Description Analyst
• Task: Extracts qualifications, required skills, and responsibilities
• Output: Structured markdown report

2. Resume & Cover Letter Agent

• Role: Professional Writer
• Task: Tailors resume to job requirements and generates personalized cover letter
• Output: Customized resume summary + professional cover letter

3. Messaging Agent

• Role: Outreach Specialist
• Task: Creates professional initial contact messages
• Output: Personalized networking/application message

App UI

🔐 Security & Privacy

• API Keys: Store in .env file (never commit to git)
• Resume Data: Processed locally, not stored permanently on external services
• Application Logs: Stored locally in CSV format

No USAJobs API results

• Check your search keyword and location are valid
• Verify USAJOBS_API_KEY is set

📊 Example Workflow

1. User searches: "Data Analyst" in "Washington, DC"
   ↓
2. USAJobs API returns 5 matching positions
   ↓
3. User selects 3 jobs to apply for
   ↓
4. For each job:
   - JD Analyst extracts requirements
   - Resume Agent tailors resume + writes cover letter
   - Messaging Agent creates outreach message
   ↓
5. All documents saved, application logged

🚀 Advanced Usage

Run Specific Agents Manually

from agent.jd_analyst import get_jd_analyst_agent, create_jd_analysis_task
from crewai import Crew

job_desc = "..."  # Job description text
agent = get_jd_analyst_agent()
task = create_jd_analysis_task(agent, job_desc)
crew = Crew(agents=[agent], tasks=[task])
result = crew.kickoff()

Customize Agent Behavior

Edit agent files in agent/ to modify:

• Agent role and backstory
• Task descriptions and expected outputs
• Model parameters (temperature, etc.)

📝 Environment Variables

GEMINI_API_KEY - Google Gemini API key for LLM

USAJOBS_API_KEY- USAJobs API key (improves results)

EMAIL_ADDRESS - Your email for tracking purposes

🤝 Contributing

Contributions are welcome! Areas for improvement:

• Additional AI agents (LinkedIn connector, salary negotiation, etc.)
• Support for other job boards
• Enhanced resume parsing
• Interview preparation agent

📄 License

MIT License - feel free to use and modify for personal use.

Acknowledgments

• CrewAI: Multi-agent orchestration framework
• Google Gemini: Advanced LLM powering the AI agents
• USAJobs: Government job listing API
• Streamlit: Web UI framework

=

Happy job hunting! 🎯 Let AI handle the application grunt work while you focus on landing your dream government role.