AI Query Microservice

A Node.js Express microservice that processes natural language queries using embeddings, vector search, and LLM-based response generation.

Features

• Query processing with semantic understanding
• Vector similarity search using Qdrant
• LLM-powered response generation with Ollama
• Redis caching for frequently asked queries
• TypeScript for type safety
• Error handling and logging

Prerequisites

• Node.js 18+ and npm
• Redis server
• Qdrant vector database
• Ollama running locally with the qwen:7b model

Setup

1. Clone the repository:

git clone <repository-url>
cd aiqueryms

2. Install dependencies:

npm install

3. Configure environment variables: Copy the .env.example file to .env and update the values:

cp .env.example .env

4. Build the project:

npm run build

5. Start the service:

npm start

For development with hot reload:

npm run dev

API Endpoints

POST /query

Process a natural language query and return an AI-generated response.

Request:

{
  "query": "What do people like about the phone?"
}

Response:

{
  "response": "Based on the reviews, people particularly appreciate..."
}

GET /health

Check the service health status.

Response:

{
  "status": "OK"
}

Environment Variables

• PORT: Server port (default: 3000)
• QDRANT_URL: Qdrant server URL
• QDRANT_COLLECTION: Qdrant collection name
• OLLAMA_API: Ollama API endpoint
• OLLAMA_MODEL: Ollama model name
• REDIS_HOST: Redis server host
• REDIS_PORT: Redis server port
• REDIS_TTL: Cache TTL in seconds
• EMBEDDING_DIMENSION: Embedding vector dimension
• TOP_K: Number of similar results to retrieve

Architecture

1. Query Processing:

   • Convert user query to embedding vector
   • Search similar vectors in Qdrant
   • Rank and filter results
   • Generate response using Ollama LLM
   • Cache results in Redis

2. Caching:

   • Redis is used to cache query-response pairs
   • Configurable TTL for cache entries
   • Improves response time for frequent queries

Error Handling

The service includes comprehensive error handling:

• Input validation
• API error handling
• Service-level error handling
• Logging with Winston

Contributing

1. Fork the repository
2. Create your feature branch
3. Commit your changes
4. Push to the branch
5. Create a new Pull Request

License

ISC