About Beam (Apache Beam ML / RunInference)
Apache Beam is an open-source unified model for batch and streaming data processing; its RunInference API enables integrating ML model inference into Beam pipelines for large-scale, streaming or batch inference.
Key Features
- RunInference API — run ML model inference inside Beam pipelines for batch and streaming workloads.
- Scales across runners (Dataflow, Flink, Spark) so inference can be executed at large scale.
- Memory and batching controls for large-model inference and LLM workflows.
- Integrates preprocessing, validation, and postprocessing steps in the same pipeline.
Use Cases & Best For
Data/ML engineers who need to run large-scale batch or streaming inference integrated with data pipelines
Teams that require pipeline-level preprocessing, validation, and inference in production
About Model Serving & APIs
Deploy and serve ML models