📄 Resume Analyzer API

An AI-powered Resume Analyzer built with FastAPI, Ollama, and Docker.

The application extracts structured candidate information from PDF resumes and compares resumes against job descriptions using a locally hosted Large Language Model (LLM).

---

✨ Features

• 📄 Analyze PDF resumes
• 🤖 AI-powered resume parsing using Ollama
• 🎯 Resume vs Job Description matching
• 📊 Structured JSON responses
• 🐳 Docker support
• ⚙️ Environment-based configuration
• ✅ Automated testing with Pytest
• 🚀 GitHub Actions CI pipeline
• 📚 Interactive Swagger API documentation

---

🛠 Tech Stack

Category	Technology
Backend	FastAPI
Language	Python 3.12
LLM	Ollama (Llama 3.2)
Containerization	Docker
Testing	Pytest
CI/CD	GitHub Actions
Documentation	Swagger / OpenAPI

---

📂 Project Structure

resume-analyzer/
│
├── config/              # Application configuration
├── data/                # Sample resumes
├── exceptions/          # Custom exception handlers
├── models/              # Pydantic response models
├── prompts/             # LLM prompts
├── services/            # Business logic
├── tests/               # Unit tests
├── uploads/             # Temporary uploaded files
├── utils/               # Utility functions
│
├── app.py               # FastAPI application
├── Dockerfile
├── docker-compose.yml
├── requirements.txt
├── README.md
└── .env.example

---

🏗 Architecture

                Client
                  │
                  ▼
           FastAPI REST API
                  │
          Upload Resume PDF
                  │
                  ▼
        Extract Text from PDF
                  │
                  ▼
       Generate LLM Prompt
                  │
                  ▼
       Ollama (Llama 3.2)
                  │
                  ▼
      Structured JSON Output
                  │
                  ▼
             API Response

---

🚀 Getting Started

1. Clone the repository

git clone https://github.com/iamanishh/resume-analyzer.git

cd resume-analyzer

---

2. Create a virtual environment

python -m venv .venv

Activate it

Windows

.venv\Scripts\activate

Linux/macOS

source .venv/bin/activate

---

3. Install dependencies

pip install -r requirements.txt

---

4. Install Ollama

Download:

https://ollama.com/download

Pull the model

ollama pull llama3.2:3b

Start Ollama

ollama serve

---

5. Configure environment

Create a .env file

MODEL_NAME=llama3.2:3b
OLLAMA_BASE_URL=http://localhost:11434

---

6. Run the application

uvicorn app:app --reload

Open

http://127.0.0.1:8000/docs

---

🐳 Running with Docker

Build the image

docker build -t resume-analyzer .

Run the container

docker run --rm -p 8000:8000 resume-analyzer

Note: When running inside Docker on Windows, configure OLLAMA_BASE_URL to use:

http://host.docker.internal:11434

---

📌 API Endpoints

Method	Endpoint	Description
GET	/	Health Check
POST	/analyze	Analyze a Resume
POST	/analyze/match	Compare Resume Against Job Description

---

📷 Screenshots

Swagger UI

Resume Analysis Response

Resume Matching Response

---

🧪 Running Tests

pytest

---

🔄 CI/CD

GitHub Actions automatically

• Installs dependencies
• Runs tests
• Builds the Docker image

on every push to the main branch.

---