Your First Inference

After install, run one end-to-end loop: seed a model → enable a flow → see predictions. This is the quickest way to confirm the runtime is working.

Tip

Each step below has a dedicated guide if you want the full detail. This page is the happy-path walkthrough.

1. Sign in

   Open the dashboard URL printed during install and sign in with your admin account.

2. Upload a model

   Go to Models and upload your ONNX model. Note its window size and feature count — they must match the input datasource the model runs on.

   See Deploying Models for formats and versioning.

3. Add an input datasource

   Go to Datasources → Input and add a source — for example an MQTT topic or an OPC-UA endpoint. Map each sensor tag to a model input channel, and set the window size to match the model.

   See Connecting Input Datasources for the full per-protocol walkthrough.

4. Create a flow

   Open Task Manager and click New Flow. Pick the input datasource, the model you uploaded, and any output datasources, then click Save. The model list only offers models whose input shape matches the input’s window size and tag count.

   See Flows for the full walkthrough.

5. Enable the flow, then press Start

   Toggle Enabled on the flow row. Enabling checks the flow again, loads the model, and connects the input — but it does not start streaming yet. Press Start in the Inference column to begin inference.

   Enable and Start are two steps

   Enable prepares the runtime (loads the model, connects the source). Start begins the data stream. Only one flow can be enabled at a time — enabling another flow turns the current one off.

6. Watch predictions

   Open the Dashboard. Within about 20 seconds (once the first window of samples arrives) the predictions card updates in real time, and the sensor cards show live readings.

Inspect results

The monitoring view shows latency (p50 / p95 / p99), throughput, error rate, and the active execution provider. See Monitoring to read the KPIs.

If something goes wrong

• The model isn’t offered for the flow, the flow won’t enable, or no predictions appear — see the Troubleshooting runbook.

Next steps

Send results out Write predictions back to a PLC or broker.

Monitor the runtime Latency, throughput, and drift.