DeepResearch

A microservices-based research platform built as a learning project for task decomposition, LangChain graphs, and vector embedding databases.

Services

Core Services

database_service

• Manages data persistence and database operations
• Provides REST API for CRUD operations on conversations, chat turns, invocations, profiles, and documents
• Uses PostgreSQL for relational data storage

embedding_service

• Handles document embedding and semantic search
• Processes documents into chunks with metadata
• Uses Qdrant vector database for similarity search
• Implements sentence transformer models for text embeddings

storage_service

• Manages file storage and retrieval
• Provides blob storage capabilities
• Integrates with database and embedding services for document processing

graph_service

• Orchestrates AI-powered research workflows
• Supports configurable LLM providers via environment settings
• Coordinates between embedding and database services

frontend_service

• Vue.js-based user interface
• Provides interactive access to research capabilities

Infrastructure Services

postgres

• PostgreSQL database for structured data storage

qdrant

• Vector database for semantic search and embeddings

minio

• Object storage service for file management

Prerequisites

• Docker and Docker Compose
• Claude API key
• Sentence transformer model specification

Environment Variables

Create a .env file in the root directory with the following variables:

# LLM provider configuration
CLAUDE_API_KEY=your_claude_api_key
OPENAI_API_KEY=your_openai_api_key
DEFAULT_LLM_MODEL=claude  # Optional: set to your preferred default model

# Embedding service
SENTENCE_TRANSFORMER_MODEL=all-MiniLM-L6-v2  # Optional; defaults to this value

# Database configuration
POSTGRES_USER=root
POSTGRES_PASSWORD=password
POSTGRES_DB=deepresearch

# MinIO configuration
MINIO_ROOT_USER=minioadmin
MINIO_ROOT_PASSWORD=minioadmin

LLM Provider Configuration

Configure your API keys and model settings in .env:

1. Set your API key: CLAUDE_API_KEY=your_claude_api_key
2. Optionally set OpenAI values: OPENAI_API_KEY and OPENAI_MODEL
3. Set default model: DEFAULT_LLM_MODEL=claude

Running the Project

Start all services:

docker-compose up

Start services in detached mode:

docker-compose up -d

Stop all services:

docker-compose down

Stop services and remove volumes:

docker-compose down -v

Service URLs

• Frontend: http://localhost:8000
• Graph Service: http://localhost:8001
• Storage Service: http://localhost:8002
• Database Service: http://localhost:8003
• Embedding Service: http://localhost:8004
• PostgreSQL: localhost:5432
• Qdrant: http://localhost:6333
• MinIO Console: http://localhost:9001

Health Checks

All services include health check endpoints at /health for monitoring service availability.