A streaming file reader built with Splittable DoFn scans an input folder for new files repeatedly, a pattern for unbounded sources in the Python SDK.

Splittable DoFn in Beam Python builds a batch file reader that processes files in parallel, and a PiSampler that estimates pi by Monte Carlo runs.

Pacing messages are added to the Beam sport activity tracker by comparing short term speed metrics against their long term counterparts.

Late droppable elements are detected by a timer in a stateful DoFn and sent to a Beam side output instead of being discarded without notice.

A stateful DoFn with Beam state and timers fixes the gRPC batch size and maximum wait time instead of leaving the bundle size to the runner.

Batching gRPC calls in a stateless DoFn so one request covers a whole bundle, cutting the time a Beam Python pipeline spends on enrichment.

The Shared class in the Beam Python SDK caches lookup data in memory for batch and streaming pipelines, with a periodic refresh for the latter.

Data augmentation in Beam Python by calling a gRPC service once per input element, running on a local Flink cluster with Kafka as the source.

A sport activity tracker in Beam Python, built first with native transforms and then with Beam SQL, showing the limits of Beam SQL in the Python SDK.

Two Beam Python pipelines compute average word length from a Kafka topic, one emitting a global average and one using a sliding time window.