Skills Extraction API

A FastAPI application that provides endpoints for searching skills using various AI models including OpenAI, Google Gemini, with vector search capabilities using Milvus/Zilliz.

Features

• Multiple skill search endpoints using different AI models:
  • Vector search using Milvus/Zilliz
  • OpenAI model-based skill extraction
  • Google Gemini AI-based skill extraction
• Response caching with GPTCache
• Docker and Docker Compose support for easy deployment
• Structured project layout following best practices

Project Structure

/your_project/
│
├── app/
│   ├── __init__.py
│   ├── main.py               # FastAPI app startup
│   │
│   ├── api/
│   │   ├── __init__.py
│   │   ├── routes.py         # API endpoints
│   │
│   ├── services/
│   │   ├── __init__.py
│   │   ├── gemini_service.py # Gemini AI service
│   │   ├── openai_service.py # OpenAI service
│   │   ├── milvus_service.py # Milvus/Zilliz service
│   │   ├── gpt_cache.py      # GPTCache integration
│   │   ├── memory_manager.py # Zilliz Cloud management
│   │
│   ├── core/
│   │   ├── __init__.py
│   │   ├── config.py         # Configuration loading
│   │   ├── logging.py        # Logging configuration
│
├── .env                      # Environment variables
├── Dockerfile                # Docker configuration
├── docker-compose.yml        # Docker Compose configuration
├── README.md                 # Project documentation
├── requirements.txt          # Python dependencies

Prerequisites

• Python 3.11+
• Docker and Docker Compose (for containerized deployment)
• API keys for OpenAI and Google Gemini
• Zilliz Cloud account (for vector database)

Environment Variables

Copy the .env.example file to .env and fill in your API keys and configuration:

# API Server Configuration
API_HOST=0.0.0.0
API_PORT=8000

# Redis Configuration
REDIS_HOST=redis
REDIS_PORT=6379
REDIS_PASSWORD=

# API Keys
OPEN_API_KEY=
GEMINI_API_KEY=

# Caching Configuration
USE_GPTCACHE=true

# Zilliz/Milvus Configuration
ZILLIZ_URI=
ZILLIZ_TOKEN=
WIPE_MILVUS_ON_START=false

# Model Configuration
GEMINI_MODEL_NAME=
OPENAI_MODEL_NAME=

Installation and Running

Using Docker Compose (Recommended)

1. Build and start the containers:

   docker-compose up -d
   

2. Access the API at http://localhost:8000

Manual Setup

1. Install dependencies:

   pip install -r requirements.txt
   

2. Run the application:

   uvicorn app.main:app --reload
   

API Endpoints

• /search?query=... - Search skills using vector embeddings in Milvus/Zilliz
• /search_openai?query=... - Extract skills using OpenAI models
• /search_gemini?query=... - Extract skills using Google Gemini AI
• /api/health - Health check endpoint

Development

To run the application in development mode:

uvicorn app.main:app --reload

This will automatically reload the application when changes are detected.