Sensor-BasedGesture Recognition
Recognizes hand and device motion gestures from IMU sensor data.
Compatible with any Nordic SoC-based device
Build AI solutions to recognize gestures and activities, detect anomalies, monitor vital signs, track and monitor assets, enable human-machine interaction in smart home and industrial systems.
Save months of tedious data gathering and training efforts with our library of ready-to-flash Edge AI models.
Experimental. These models are designed for rapid prototyping. Before deploying to production, we recommend completing your own testing.
Sensor-BasedRecognizes hand and device motion gestures from IMU sensor data.
Keyword SpottingVoice commands for media playback — play, stop, next, loud, quiet.
Wake WordTrigger your device hands-free with the “Okay Nordic” wake phrase.
Sound EventsDetects snoring sounds in incoming audio in real time.
Keyword SpottingVoice commands for door locks and access management — access, alarm, deny, grant, help, lock, status, visitor.
Keyword SpottingVoice commands for automated cleaning devices — clean, continue, dock, mop, pause, silent, spot, stop, turbo.
Choose the use case that fits
One optimized archive, ready to flash
No setup, no training
Build ultra-low-power, on-device AI optimized for Nordic hardware.
Designed for ultra-low-power edge devices
Build and deploy across hardware targets
No ML expertise required
Transform raw sensor streams into high-quality training data using built-in preprocessing tools designed for edge AI.
Split continuous sensor data into time windows to capture meaningful patterns
Automatically generate signal features that help the model learn more efficiently
Automatically identify and retain only the most relevant features for the model
Smaller models. Faster inference. Better accuracy.
Understand how your data & model performs and identify opportunities for improvement after training.
Visualize model performance across multiple metrics to quickly assess overall model quality

Understand which features contribute most to model predictions and identify redundant inputs

Analyze prediction accuracy across classes and identify common misclassifications

Identify the most important features in your dataset based on statistical significance

Create production-ready wake word & keyword spotting models in just a few steps


Define how to wake up and control your device
Automatic training (1-3 hours)
Test the model directly in your browser
Download model and run inference on device