Skip to content

Run AI applications in production

Managed Machine Learning platform

Ingest from any data source

  • Read data from any cloud environment or on-premise
  • Validate data for missing features
  • Take into account data drift using anomaly detection

Process Data

  • Perform scalable preprocessing of the data to handle any training dataset size
  • Use TensorFlowTransform (TFT) to define the preprocessing operations
  • Avoid training/serving skew

Train and deploy models anywhere

  • Train models on any scale
  • Validate models automatically before deploying them
  • Deploy models in any cloud or on the edge

IManaged Machine Learning platform?

Pipeline version control

ewfwef

Cloud and local implementation