Скачать или смотреть DeltaStream for Real Time OCI Streaming for Kafka Pipelines

In this video, see how DeltaStream works with OCI Streaming for Kafka to build real-time data pipelines without operational overhead.

Many Oracle users already rely on event-driven architectures to capture customer activity, device telemetry, financial events, and internal operational data. OCI Streaming for Kafka is a natural home for this event traffic, but turning raw Kafka events into usable, downstream-ready data often requires standing up and maintaining Flink clusters, connectors, and complex streaming infrastructure.

DeltaStream removes that complexity.

In this demo, we show how teams can:

Connect DeltaStream to OCI Streaming for Kafka using Kafka-compatible APIs

Browse and preview Kafka topics in JSON or table-like form

Define streaming schemas using familiar SQL

Run interactive SQL queries to filter, join, and explore live data

Create transformed Kafka topics with a single SQL statement

Monitor throughput and latency without managing streaming compute

With DeltaStream, streaming data can be treated like a database table. Users write SQL, and DeltaStream automatically provisions and manages the underlying streaming compute, applies the transformation logic, and publishes the results back to Kafka as first-class topics.

The demo walks through creating a stream from a Kafka topic (for example, page view events), filtering for specific business events such as purchases, and publishing the transformed results back into OCI Streaming for Kafka, all without owning or operating Flink clusters.

While this video focuses on OCI Streaming for Kafka, DeltaStream can also connect to CDC streams, legacy queues, cloud storage, and other streaming systems. The same SQL-based approach can be used to enrich and join data across sources and write results back to Kafka or downstream platforms like Iceberg, Databricks, and Snowflake.

OCI Streaming for Kafka provides the event backbone. DeltaStream turns those events into clean, actionable, real-time data, without the operational burden.