Quick Start

This guide will help you get started with Lattice—whether you’re an end user looking to run the app or a developer integrating with the API.

App User Install and run Lattice locally or on your private cloud

API Developer Integrate with the Lattice API for custom workflows

Get Access

Lattice API Access: $90 one-time fee

Includes:

• Full access to Lattice app and API
• Run locally on your laptop
• Host on your private cloud
• No recurring subscription fees
• All future updates included

Visit latticelab.io to purchase and receive access to the private GitHub repository.

---

Installation

Prerequisites

Before you begin, ensure you have:

• Git — For cloning the repository
• Docker & Docker Compose — For PostgreSQL database
• Python 3.11+ — For the backend server
• Node.js 18+ — For the frontend application
• An API key from at least one AI provider (Anthropic, OpenAI, Google, or Ollama)

Step 1: Clone the Repository

After purchasing, you’ll receive access to the private GitHub repository.

git clone https://github.com/navam-io/lattice.git
cd lattice

Step 2: Start the Database

Lattice uses PostgreSQL with pgvector for vector embeddings. Start the database using Docker Compose:

docker compose up -d

This starts a PostgreSQL 16 container with:

• Container name: lattice-postgres
• Database: lattice
• User/Password: lattice / lattice_dev
• Port: 5432

Verify the container is running:

docker compose ps

Step 3: Set Up the Backend

Using pip

cd backend

# Create virtual environment
python3 -m venv .venv
source .venv/bin/activate

# Install dependencies
pip install -e ".[dev]"

Using uv (faster)

cd backend

# Install uv if not already installed
curl -LsSf https://astral.sh/uv/install.sh | sh

# Sync dependencies
uv sync

Step 4: Configure Environment Variables

Create a .env file in the backend directory:

cd backend
touch .env

Add your configuration:

# Required
ANTHROPIC_API_KEY=sk-ant-...

# Optional providers
OPENAI_API_KEY=sk-...
GOOGLE_API_KEY=AIza...

# Database (default Docker Compose settings)
DATABASE_URL=postgresql+asyncpg://lattice:lattice_dev@localhost:5432/lattice

# Application
ENVIRONMENT=development
DEBUG=true

Tip

At minimum, you need one AI provider API key. Anthropic is recommended for the best experience with Lattice’s reasoning capabilities.

Step 5: Run Database Migrations

Apply the database schema:

cd backend
alembic upgrade head

Step 6: Start the Backend Server

cd backend
uvicorn main:app --reload --port 8000

The API is now available at http://localhost:8000.

Verify it’s running:

curl http://localhost:8000/api/health

Step 7: Set Up the Frontend

In a new terminal:

cd frontend
npm install

Step 8: Start the Frontend

Development

npm run dev

Access the app at http://localhost:3000 with hot reload enabled.

Production

# Build for production
npm run build

# Start production server
npm start

Optimized build for production deployment.

Note

For production deployments, ensure you set ENVIRONMENT=production and configure proper security settings.

---

Using the App

Once installation is complete, navigate to http://localhost:3000 in your browser.

1. Create a Workspace

   Click the workspace selector and choose + New Workspace. Give it a descriptive name that reflects your research focus, like “Claude Sonnet Evaluation” or “RAG Pipeline Comparison”.

2. Add Your First Sources

   Click the + button in the Sources panel to add knowledge sources:

   • URL — Paste documentation links, pricing pages, or blog posts
   • PDF — Upload research papers, model cards, or internal docs
   • GitHub — Connect repositories for code analysis

   Tip

   Start with 3-5 key sources. You can always add more as your research deepens.

3. Start Chatting in the Lab

   Once your sources are indexed, ask questions in the Lab panel:

   What are the key differences between Claude Sonnet and GPT-4
   for production RAG applications?

   Lattice will search your sources and provide grounded responses with citations.

4. Save Artifacts to Studio

   When the AI generates useful tables, comparisons, or analyses, click Save to Studio to preserve them for later reference.

Using Blueprints for Quick Setup

Don’t want to curate sources manually? Use a Blueprint to get started instantly:

1. Navigate to Blueprints in the navigation
2. Browse available blueprints by vendor or use case
3. Click Apply to import curated sources, scenarios, and stacks
4. Start researching immediately with pre-configured knowledge

Note

Blueprints include official documentation from vendors like Anthropic, OpenAI, AWS, and NVIDIA—automatically kept up to date.

---

API Access

Lattice API is designed for local self-hosted deployment and does not require authentication. All endpoints are accessible without credentials.

API Endpoint

Endpoint	Description
http://localhost:8000	Local development

Interactive Documentation

Once the backend is running, access interactive API documentation at:

• Swagger UI: http://localhost:8000/docs
• OpenAPI Spec: http://localhost:8000/openapi.json

LLM Provider Keys

Lattice requires API keys for AI providers (Anthropic, OpenAI, etc.) which are configured via:

• Settings Panel — Configure keys through the app UI
• API Endpoint — Programmatically manage keys via /api/api-keys

Note

LLM provider keys are stored encrypted in the database and are separate from API authentication.

Core API Examples

List Workspaces

curl http://localhost:8000/api/workspaces

Create Workspace

curl -X POST http://localhost:8000/api/workspaces \
  -H "Content-Type: application/json" \
  -d '{
    "name": "My Research Project",
    "description": "Evaluating LLMs for production"
  }'

Add Source

curl -X POST http://localhost:8000/api/sources \
  -H "Content-Type: application/json" \
  -d '{
    "workspace_id": "ws_123",
    "type": "url",
    "url": "https://docs.anthropic.com/en/docs"
  }'

Chat Query

curl -X POST http://localhost:8000/api/messages \
  -H "Content-Type: application/json" \
  -d '{
    "workspace_id": "ws_123",
    "content": "Compare Claude 3.5 Sonnet and GPT-4 for RAG"
  }'

Health Check

Verify the API is running:

# Basic health check
curl http://localhost:8000/api/health

# Readiness probe (includes database check)
curl http://localhost:8000/api/health/ready

Tip

See the full API Reference for complete endpoint documentation.

---

Troubleshooting

Database Connection Issues

# Check if PostgreSQL container is running
docker compose ps

# View container logs
docker compose logs postgres

# Restart the database
docker compose restart postgres

Port Already in Use

# Kill processes on specific ports
lsof -ti:3000 | xargs kill -9  # Frontend
lsof -ti:8000 | xargs kill -9  # Backend

Missing Dependencies

# Backend - reinstall dependencies
cd backend && pip install -e ".[dev]" --force-reinstall

# Frontend - clean install
cd frontend && rm -rf node_modules package-lock.json && npm install

---

Next Steps

Now that you’re set up:

• Core Concepts — Understand workspaces, sources, scenarios, and stacks
• Features Overview — Explore all Lattice capabilities
• Evaluate AI Models — Complete your first model evaluation
• API Reference — Full API documentation for developers