Quickstart

Get TrustGraph running and see it work with your own eyes. By the end of this guide you’ll have a working instance, loaded data, queried a knowledge graph, and seen explainability in action.

What you’re going to do

  1. Launch TrustGraph using Docker Compose or Podman Compose on a single machine — a laptop, workstation, or cloud instance
  2. Load documents and process them into a knowledge graph
  3. Query with Graph RAG and see how answers trace back to sources
  4. Use ontologies to control what knowledge gets extracted
  5. Run SPARQL queries against structured knowledge
  6. Explore the graph visually in the Context Graph viewer

The whole thing takes under 30 minutes.

What you’ll need

  • A machine with 12GB+ RAM and 8 CPUs available for TrustGraph — a 16GB laptop will work if nothing else heavy is running
  • Docker Engine or Podman installed
  • Python 3 for CLI tools
  • Access to an LLM — a cloud API key (OpenAI, Anthropic, Google VertexAI, AWS Bedrock) or a local model via Ollama
  • Basic command-line familiarity

If Docker Compose on a single machine isn’t right for your situation, see the deployment options for Kubernetes, cloud platforms, and other approaches. This quickstart focuses on the simplest path to get you running.

Next

Prepare the configuration — generate your deployment config.

Table of contents