Multi-Agent Hiring System

This project is a multi-agent system designed to streamline the hiring process by leveraging AI and automation. It uses the Ollama-hosted Gemma3 model for natural language processing tasks. The system automates job description summarization, CV extraction, candidate selection, and email notifications for shortlisted candidates.

---

Table of Contents

1. Overview
2. Agents in the System
3. Usage Instructions
4. Example Execution
5. Database Management
6. Text Files
7. Dependencies
8. Dataset
9. License

---

Overview

The Multi-Agent Hiring System automates the recruitment workflow by breaking it into modular tasks handled by specialized agents. Each agent performs a specific task, ensuring scalability, modularity, and maintainability. The system integrates with an SQLite database for persistent storage and supports both real and simulated email notifications.

---

Agents in the System

1. JobDescriptionSummarizerAgent:

   • Summarizes the provided job description into key elements (e.g., skills, experience, qualifications).
   • Writes the summary to job_description_summary.txt.

2. CandidateExtractorAgent:

   • Extracts and summarizes key information (e.g., skills, experience) from CVs in PDF format.
   • Writes the summaries to candidate_data.txt.

3. HiringManagerAgent:

   • Matches candidates' CV summaries with the job description summary.
   • Selects the best candidates based on predefined criteria.
   • Writes the selection response to hiring_manager_response.txt.

4. CandidateDataRetrieverAgent:

   • Extracts specific details ( name, candidate ID, email) for selected candidates.
   • Prepares data for email notifications.

---

Usage Instructions

Prerequisites

• Python 3.8 or higher installed.
• Install Ollama and host the Gemma3 model locally.
• Ensure all required libraries are installed (see Dependencies).

Steps to Run the System

1. Clone the Repository:

   git clone https://github.com/Mohammed-Saajid/MultiAgentHiringSystem.git
   cd MultiAgentHiringSystem

2. Install Dependencies:

   pip install -r requirements.txt

3. Run the Application: Execute the app.py script in the src folder with the following parameters:

   python src/app.py --job_title "Job Title" --job_description "Job Description" --cv_directory "path/to/cv/folder" --email_test 0

   • --job_title: The title of the job (e.g., "Software Engineer").
   • --job_description: The full job description text.
   • --cv_directory: Path to the folder containing CVs in PDF format.
   • --email_test: Set to 1 to send real emails or 0 to simulate sending email invitations.

---

Example Execution

Using example.bat

1. Navigate to the example folder:

   cd example

2. Run the batch file:

   example.bat

   This will execute the system with 3 sample CVs located in the test folder.

Using example.py

1. Navigate to the example folder:

   cd example

2. Run the script:

   python example.py

   This will process the entire dataset of job descriptions and CVs.

---

Database Management

• The system uses an SQLite database for long-term memory storage.
• To reset the database, run the reset_db.py script:

  python src/reset_db.py

---

Text Files

The following text files are used to store intermediate outputs:

• candidate_data.txt: Contains summaries of CVs processed by the CandidateExtractorAgent.
• hiring_manager_response.txt: Contains responses from the HiringManagerAgent regarding candidate selection.
• job_description_summary.txt: Contains the summarized job description generated by the JobDescriptionSummarizerAgent.

---

Dependencies

Install the required libraries using requirements.txt:

pip install -r requirements.txt

Dataset

The dataset utilized in this project is provided by Hack the Future: A Gen AI Sprint Powered by Data, a hackathon hosted collaboratively by Accenture and GeeksforGeeks.

License

This project is licensed under the Apache License. See the LICENSE file for details.