Add blog: Streaming Real-Time Features to Lakebase with the Native Sink

[muralitalluri-db]

Adds a self-serve example under 2026-09-rtm-native-lakebase-sink/ that uses Spark RTM to compute features and ingest them to Databricks Lakebase via the native Lakebase sink — for a real-time personalization use case.

What it does:

• Reads video engagement events (watches/likes) from Kafka
• Computes each user's last-6 watched / last-3 liked with transformWithState (single ValueState)
• Writes one feature row per user to a UC-registered Lakebase table via .toTable() (native sink) — no ForeachWriter, no offline→online sync

Contents:

• ingest-source-data/ — data generator, Kafka topic admin, Delta→Kafka replay
• sss-rtm-to-lakebase/ — Lakebase table creation + the RTM feature pipeline
• README.md — end-to-end self-serve instructions (workspace-specific values are placeholders)

Requires DBR 18 LTS.

[dichlorodiphen]

LGTM (david-young_data)