Try it on a laptop

A demo, not a deployment

Production runs on the NVIDIA Jetson only. To install a box, follow Jetson Deployment and the Offline Bundle Install. This page starts a demo stack on a laptop so you can explore the dashboard.

The demo runs the same three services as the box — inference, backend, and dashboard — with inference on the CPU and a small test model, so the stack starts healthy without any edge hardware.

Prerequisites

• Docker Engine with the Docker Compose v2 plugin
• A checkout of the X-Edge AI source repository
• Internet access on the first run, while Docker builds the images

1. Start the stack

From the root of the repository:

./start-aiboard.sh

On the first run, the script writes a fresh JWT signing secret into a local .env, then starts the stack and waits for it to become healthy. Building the images the first time takes several minutes.

2. Create an admin account

docker exec aiboard-backend-phase1.1 \
  dotnet AiBoard.Orchestration.dll seed-admin \
  --username admin --password '<choose-a-password>' --role admin

3. Open the dashboard

Open http://localhost (port 3000 also works) and sign in as admin.

4. Next steps

• First Inference — load a model and see live predictions
• Connect an input datasource — replay a CSV file, the fastest input to try
• Docker Lab Walkthrough — add OPC UA and MQTT simulators

To stop the demo, run docker compose down from the repository root.