AI@Edge

/IOTCONNECT™
AI@Edge

AI@Edge brings machine-learning models to your IoT devices so they can make smart decisions right where the data is produced — without waiting on the cloud. /IOTCONNECT™ gives you everything you need: a shared Model Hub to manage and distribute models, SageMaker-powered model conversion, and ST retraining so your edge models keep improving over time.

What is AI@Edge?

AI@Edge means running artificial-intelligence models directly on your IoT devices (the “edge”) instead of only in the cloud. The device can detect anomalies, classify sensor patterns, or make predictions in real time — even with limited or no connectivity.

/IOTCONNECT™ supports the full lifecycle of an edge model through three building blocks:

Model Hub: A central library where models are created, versioned, and shared. Platform admins publish global models that every account can use, and companies can manage their own models.

Conversation with SageMaker: How Amazon SageMaker converts and prepares a model so it is ready to deploy, without you having to upload files manually.

Retrain with ST: How device data is sent to the ST AIoT Craft service to retrain a model, so your edge AI keeps getting better with real-world data.

Use the links above (or the sidebar) to dive into each topic.

ⓘ NOTE

New to edge AI? Start with the Model Hub to understand how models are organized and shared, then read Conversation with SageMaker and Retrain with ST to see how models are prepared and improved.