Real-Time Streaming

Data as it happens. Kafka, event time, windows, watermarks, stateful processing and delivery guarantees.

• Stream Processing: Kafka, Windows & Exactly-Once — Processing data as it happens. Learn event streaming and Apache Kafka (topics, partitions, offsets, producers, consumer groups, retention, replication), event time vs processing time, windows (tumbling, sliding/hopping, session), watermarks and late data, stateful processing and checkpoints (Flink, Spark Structured Streaming, Kafka Streams), delivery semantics (at-most-once, at-least-once, exactly-once), stream–table joins, CDC streams, schema registries, and lambda vs kappa architectures — with Python simulating tumbling windows with a watermark that drops late events.