ATS Resume Tracker

Overview

The ATS Resume Tracker is a full-stack web application that provides a comprehensive, AI-powered analysis of resumes against specific job descriptions (JDs). It utilizes Google Gemini API for contextual understanding, LangChain for orchestration, and MongoDB with vector search to implement Retrieval-Augmented Generation (RAG). This ensures deep semantic matching between resumes and job descriptions.

The system offers a professional evaluation of uploaded PDF resumes, computes a percentage match, generates a strengths and weaknesses analysis, and enhances accessibility with audio playback and YouTube video generation. The results can be viewed on the site, emailed to the user, or downloaded as a PDF report.

---

Project Demonstration:

Recording.2025-06-21.205535.mp4

Application Interface:

Lgin and Registeration

Upload and Analyze Resume:

Evaluation Result Page:

Emailed Report Screenshot:

Youtube vedio recommendation:

Project Features

• Resume Upload and Parsing: Accepts resumes in PDF format and extracts structured content such as skills, education, experience, etc.
• Contextual JD Matching: Uses LangChain + Gemini API + MongoDB vector index to retrieve and compare relevant information from the resume and job description.
• Strengths and Weaknesses Evaluation: AI-generated summary that highlights what a candidate excels in and areas needing improvement.
• Percentage Match Score: Reflects how closely the resume aligns with the job description based on AI-driven semantic similarity.
• Audio Playback: The result section includes a "Play Audio" button that reads the summary aloud using text-to-speech.
• PDF Download: The complete evaluation can be downloaded by the user as a professionally formatted PDF.
• YouTube Integration: Based on the evaluation, the app uses the YouTube API to recommend personalized videos (e.g., resume tips, upskilling tutorials) directly within the UI.
• Email Notification: The final result is emailed to the user for easy reference.
• Modern Web Interface: Built using Flask, HTML, CSS, and JavaScript to ensure responsive and user-friendly design.

---

Tech Stack

Layer	Technology
Backend	Python, Flask
Frontend	HTML, CSS, JavaScript
Gen AI model	Gemini 2.5 Flash (Google Generative AI) for reasoning & generation; LangChain for orchestration
Embeddings	text-embedding-004 (Google Generative AI) for vector representation
Database	MongoDB Atlas with vector search
Audio	Client-side audio playback using the Web Speech API (SpeechSynthesisUtterance)
Video Suggestion	YouTube Data API
Email Delivery	SMTP (Gmail)
PDF Generation	Built-in browser print dialog via JavaScript’s window.print()

Prerequisites

• Python 3.x
• MongoDB Atlas account with vector indexing enabled
• Gemini API access (Google Generative AI)
• Gmail account with app password (for SMTP)
• YouTube API key

---

Repository Setup

Clone the Repository

git clone https://github.com/sahilchalke0001/resume
cd resume

Create a Virtual Environment

python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate

Install Dependencies

pip install -r requirements.txt

Environment Configuration

Create a .env file in the root directory and add the following keys:

FLASK_SECRET_KEY=4ddcdd430dc9d6f3848cc171d0c54358f2aa9e3f2a8fcc27167353105a4b610d

# Google Generative AI / Gemini API
GEMINI_API_KEY=your_api_key_here
GOOGLE_API_KEY=your_google_api_key_here

# Gmail for SMTP
SENDER_EMAIL=your_email@gmail.com
EMAIL_PASSWORD=your_gmail_app_password

# YouTube API (for video suggestions)
YOUTUBE_API_KEY=your_youtube_api_key

# MongoDB RAG Setup
MONGO_URI=mongodb+srv://<username>:<password>@cluster0.mongodb.net/db?retryWrites=true&w=majority&appName=Cluster0
MONGO_DB_NAME=db
MONGO_KB_COLLECTION=knowledge_chunks
MONGO_VECTOR_INDEX=default

---

Running the Application

python app.py

Then visit:

http://127.0.0.1:5000/

---

How to Use

1. Open the web interface and upload your resume in PDF format.

2. Paste or write your target job description.

3. Click the Analyze button.

4. The application will:

   • Parse the resume
   • Contextually match it with the job description using RAG
   • Generate a percentage match
   • Summarize key strengths and weaknesses
   • Provide video suggestions
   • Offer audio narration of the result
   • Allow download of the PDF
   • Email the complete report

---

License

This project is licensed under the MIT License.

---

Contact

For queries, feedback, or collaboration opportunities, please contact: kalepranav2608@gmail.com

---