Job Agent Automation Project

Overview

This project is an automated job application tool that integrates a browser controller, PDF reader, and language models to help users find and apply for jobs efficiently. It reads a user's CV, searches for relevant job postings, evaluates job fit scores, and generates personalized cover letters for applications.

Features

• Read CV: Extracts text from the user's CV (PDF format) to use as context.
• Job Search: Finds job postings for machine learning and data science positions in Germany (Can be changed as per requirements).
• Job Evaluation: Rates job postings based on fit with the user's profile.
• Cover Letter Generation: Automatically generates cover letters for applications.
• File Handling: Saves job postings and reads them back from a CSV file.
• Logging: Provides detailed logs for debugging and tracking operations.

Project Structure

Job Agent
|-- Lambda functions
    |-- extract-cv-text.py
    |-- parse-cv.py
    |-- extract-job-keywords.py
    |-- fetch-jobs.py
    |-- job-scoring.py
    |-- store-jobs-matches-to-s3.py
|-- Job Agent.log
|-- jobs.csv
|-- jobAgent.py
|-- .env

Lambda Functions

📄 extract-cv-text.py

• This Lambda function is triggered by an S3 event when a CV PDF is uploaded. It extracts the raw text from the PDF using PyPDF2, then invokes the parse-cv Lambda function by passing the extracted text.
• This serves as the entry point for processing candidate resumes.

🧠 parse-cv.py

• This Lambda function receives raw resume text, parses it using the Gemini LLM to extract structured information (e.g., name, email, skills, education, experience), and stores it in S3.
• It then triggers a chain of Lambda functions to extract job-related keywords, fetch relevant job listings, score them based on CV fit, and finally store the top results in another S3 bucket.

🔍 extract-job-keywords.py

• This Lambda function takes a structured CV JSON as input and uses Gemini LLM to extract the top 3 most relevant job field keywords.
• It filters out tools, libraries, and generic terms, returning only the core areas of expertise suitable for matching relevant job listings.

📄 fetch-jobs.py

• This Lambda function queries the Adzuna Jobs API using a list of relevant job field keywords extracted from a candidate's profile.
• It retrieves job postings for each keyword and returns a combined list of matching job results.

🎯 job-scoring.py

• This Lambda function evaluates how well each job posting matches a candidate's profile using the Gemini LLM.
• It generates a relevance score (0–100) and an explanation for each job by comparing the candidate's skills, education, and experience with the job description.
• The scores are returned in JSON format.

💾 store-job-matches-to-s3.py

• This Lambda function stores scored job matches into an Amazon S3 bucket as a CSV file.
• Each user has a dedicated folder (based on their email), and job entries are organized in tabular format. If a job with the same ID already exists, it is automatically updated with the new score and details.

APIs

This project defines multiple secure API endpoints.

Job-CV Matching API

• This API securely sends parsed CV and job data to an AWS Lambda function, which uses prompt engineering with a Large Language Model (LLM) (Gemini) to evaluate how well a CV matches each job.

  🚀 How to Use This API

  URL: https://aydopiql03.execute-api.eu-central-1.amazonaws.com/development-env

  Make a POST request to the API with the following JSON payload:

    {
    "cv": {
        "Full Name": "YOUR NAME",
        "Email": "name@domain.com",
        "Skills": "Some skills like C, Python, R, ...",
        "Education": [
        {
            "Degree": "Master’s in xxx, xxxx",
            "Institution": "xxxx, xxxx",
            "Duration": "xxxx-xx"
        },
        {
            "Degree": "Bachelor of xxxx, xxxx ",
            "Institution": "xxxx, xxxx",
            "Duration": "xxxx-xx"
        }
        ],
        "Work Experience": [
        {
            "Title": "xxxx",
            "Company": "xxxx",
            "Duration": "xxx - Present"
        },
        ]
    },
    "jobs": [
        {
        "id": "xxxx",
        "title": "xxxx",
        "company": {
            "display_name": "xxxx"
        },
        "description": "xxxx",
        "location": {
            "display_name": "xxxx"
        },
        }
    ]
    }
  

  Response from the API:

    {
        "statusCode": 200,
        "body": "{\"jobs\": [{...}]}"
    }
  

  Each job in the response has the following structure:

    {
        "id": "xxxx",
        "title": "xxxx",
        "company": {
            "display_name": "xxxx"
        },
        "description": "xxxx",
        "location": {
            "display_name": "xxxx"
        },
        "score": {
            "score": "0-100",
            "reason": "xxxx"
        }
    }
  

Key Components

• jobAgent.py: The main entry point of the project.
• Job Agent.log: Log file to track application events.
• jobs.csv: File where job postings are saved.
• .env: Environment file to store sensitive information such as API keys and file paths.
• app.py: Python file to connect with locally hosted deepseek-r1 model.

Requirements

• Python 3.11+ (browser_use requirement)
• Dependencies listed in requirements.txt

Key Libraries Used

• PyPDF2: For reading and extracting text from PDF files.
• asyncio: For asynchronous task handling.
• dotenv: For managing environment variables.
• pydantic: For data validation.
• csv: For file operations.
• logging: For detailed logging.

Setup Instructions

1. Clone the Repository

git clone https://github.com/hemantsirsat/Job-Agent.git
cd Job-Agent

2. Install Dependencies

pip install -r requirements.txt

3. Set Up Environment Variables

Create a .env file with the following variables:

CV_PATH=<path_to_your_cv_pdf>
GEMINI_API=<your_gemini_api_key>

4. Run the Project

python jobAgent.py

Logging

Logs are saved in Job Agent.log. They include detailed information about operations and errors.

Example Log Output

2025-01-29 Tuesday 10:30:21 INFO:CV path = /path/to/cv.pdf
2025-01-29 Tuesday 10:31:05 INFO:Saved job to file successfully.

Contributions

Contributions are welcome! Please fork the repository and submit a pull request.