AI Job Application Tracker

AI-powered backend application that helps users manage job applications and analyze job descriptions using Large Language Models (LLMs). Built with FastAPI, PostgreSQL, SQLAlchemy, JWT Authentication, and Groq's Llama 3.1 model.

Features

• User Registration and Login with JWT Authentication
• Secure REST APIs with Protected Endpoints
• Job Application Tracking and Management
• AI-Powered Job Description Analysis
• Extraction of Required Skills, Experience Level, Responsibilities, and Salary Insights
• PostgreSQL Database Integration
• SQLAlchemy ORM for Database Operations
• Database Versioning with Alembic Migrations
• Rate Limiting using SlowAPI
• Dockerized Application Setup
• Automated Testing with Pytest
• Clean Layered Architecture (Routes, Services, Models, Schemas)
• React Frontend Integration Planned
• Replacing Llama 3.8B model with GPT OSS 20B *Adding Rate Limiting

System Architecture

The application follows a layered architecture:

• Client Layer (React Frontend, Swagger UI, Postman)
• FastAPI API Layer
• Authentication & Dependency Layer
• Business Logic Service Layer
• PostgreSQL Persistence Layer
• Groq AI Integration Layer
• Dockerized Deployment Layer

Tech Stack

Layer	Technology
Backend	FastAPI
Database	PostgreSQL
ORM	SQLAlchemy
Authentication	JWT (HTTPBearer)
AI Model	Groq (Llama 3.1 8B)
Migrations	Alembic
Rate Limiting	SlowAPI
Testing	Pytest
Containerization	Docker
Deployment	Railway

Project Structure

ai_analyzer/
│
├── app/
│   ├── core/
│   ├── dependencies/
│   ├── models/
│   ├── routes/
│   ├── schemas/
│   ├── services/
│   └── utils/
│
├── migrations/
├── tests/
├── Dockerfile
├── docker-compose.yml
├── requirements.txt
├── alembic.ini
└── README.md

AUTH FLOW

Client │ POST /login │ JWT Token │ Authorization: Bearer │
Protected Endpoint

Setup

1. Clone the Repository

git clone https://github.com/Farru049/ai_analyzer.git
cd ai_analyzer

2. Create Virtual Environment

python -m venv myenv

3. Activate Virtual Environment

Windows:

myenv\Scripts\activate

Linux / macOS:

source myenv/bin/activate

4. Install Dependencies

pip install -r requirements.txt

5. Create PostgreSQL Database

analyzer_db

6. Configure Environment Variables

Create a .env file:

DATABASE_URL=postgresql://username:password@localhost:5432/analyzer_db

SECRET_KEY=your_secret_key

ALGORITHM=HS256

ACCESS_TOKEN_EXPIRE_MINUTES=30

GROQ_API_KEY=your_groq_api_key

7. Run Database Migrations

alembic upgrade head

8. Start the Application

uvicorn app.main:app --reload

Application will be available at:

http://127.0.0.1:8000

Swagger Documentation:

http://127.0.0.1:8000/docs

Running with Docker

Build and run:

docker-compose up --build

Running Tests

pytest

API Endpoints

Authentication

Method	Endpoint	Description
POST	/register	Register User
POST	/login	Login User

Applications

Method	Endpoint	Description
GET	/applications	Get All Applications
POST	/applications	Create Application
GET	/applications/{id}	Get Application
PUT	/applications/{id}	Update Application
DELETE	/applications/{id}	Delete Application

AI

Method	Endpoint	Description
POST	/ai/analyze	Analyze Job Description

AI Analysis Workflow

1. User submits a job description.
2. FastAPI validates and authenticates the request.
3. AI Service constructs a prompt.
4. Prompt is sent to Groq's Llama 3.1 model.
5. AI-generated insights are returned.
6. Analysis can be stored alongside the application record.

Database Schema

Users

• id
• email
• hashed_password
• created_at

Applications

• id
• user_id
• company_name
• role
• status
• description
• analysis
• notes
• applied_at

Relationship:

User (1) ------ (N) Applications

Deployment

• Dockerized Application
• Railway Deployment
• PostgreSQL Database
• Groq Cloud AI Service

Future Enhancements

• React Frontend Integration
• Redis Caching
• Token Blacklisting with Redis
• Async SQLAlchemy
• CI/CD using GitHub Actions
• Prometheus Monitoring
• Role-Based Access Control (RBAC)

Author

Mohammad Farhaan Ali

GitHub: https://github.com/Farru049