Installation Guide

⚠️ Pre-requisites

Before proceeding, ensure your system meets these requirements:

✅ Docker Engine 24.0+ and Docker Compose 2.20+
✅ 8GB RAM minimum (16GB recommended for local models)
✅ 25GB available disk space
✅ NVIDIA GPU with drivers v535+ (optional for GPU acceleration)
✅ Unix-based OS (Linux/macOS/WSL2 on Windows)

Also ensure to follow the prerequisites tutorial Requirements

🚀 Quick Installation

1. Create docker-compose.yaml

Create a file with this content:

services:
  ollama:
    image: manuc1k/demo-inside-the-lab-ollama:latest
    # deploy:
    #   resources:
    #     reservations:
    #       devices:
    #         - driver: nvidia
    #           count: all
    #           capabilities: [gpu]
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:11434/"]
      interval: 30s
      timeout: 20s
      retries: 5
      start_period: 60s

  database:
    image: manuc1k/demo-inside-the-lab-database:latest
    healthcheck:
      test: echo 'db.runCommand("ping").ok' | mongosh localhost:27017/test --quiet
      interval: 20s
      timeout: 10s
      retries: 5
      start_period: 60s

  qdrant:
    image: manuc1k/demo-inside-the-lab-qdrant:latest
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:6333/healthz"]
      interval: 30s
      timeout: 10s
      retries: 10
      start_period: 60s

  backend:
    image: manuc1k/demo-inside-the-lab-backend:latest
    depends_on:
      database:
        condition: service_healthy
      qdrant:
        condition: service_healthy
      ollama:
        condition: service_healthy

  frontend:
    image: manuc1k/demo-inside-the-lab-frontend:latest
    ports:
      - "3000:3000"
    depends_on:
      - backend

2. Start the Application

# Start all services with GPU support
docker compose up -d --pull always

# Monitor initialization (wait for all services healthy)
docker compose logs -f

3. Access the Application

Once all services show as healthy (2-3 minutes):

🌐 Open http://localhost:3000 in your browser

🔍 Verification

Check all services are running:

docker compose ps

Expected output:

NAME                SERVICE             STATUS              PORTS
chatty-ollama-1     ollama              running             
chatty-database-1   database            healthy             27017/tcp
chatty-qdrant-1     qdrant              healthy             6333/tcp, 6334/tcp
chatty-backend-1    backend             running             
chatty-frontend-1   frontend            running             0.0.0.0:3000->3000/tcp

🛠️ Basic Configuration

Set API Keys (Optional)

Edit the docker-compose.yaml and add under backend.environment:

- OPENAI_KEY=your_openai_key
- MISTRAL_KEY=your_mistral_key
- GEMINI_KEY=your_gemini_key

🖥️ GPU Acceleration (Recommended)

⚠️ Before using:
Configuración de NVIDIA para Docker

(Must have NVIDIA drivers v535+ and Docker GPU support)

Critical for optimal performance
Using GPU acceleration reduces Ollama's response time by 5-8x. For NVIDIA GPUs:

ollama:
    image: manuc1k/demo-inside-the-lab-ollama:latest
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: all  # Utilizes all available GPUs
              capabilities: [gpu]  # Requires NVIDIA Container Toolkit