1/ @apache_pulsar 5.0 is out, and it's the new LTS. The milestone: no more partitions. With Scalable Topics, you never have to answer "how many partitions?" again.

Data Streaming Summit (organized by @streamnativeio) back. October 7–8, San Francisco. 🎉 Same hardcore, community-first conference we've always run. Same room full of practitioners running real-time data infrastructure at real scale. Same bias toward production numbers over keynote theater. What's new: we're adding two tracks for the era we're actually living in. → Data Streaming Engines — the foundation. Pulsar, Kafka, Ursa, Flink, diskless topics, leaderless logs, lakehouse-native storage. → Data Streaming for Agents — the bridge. Real-time context, durable memory, event-driven tool orchestration, agent observability. → Agent Harness, Runtime & Governance — the agentic layer. Managed Agents, Agent Executor, Deep Agents, and many in-house runtimes enterprises are quietly building. Here's the thing: the agent community is running into problems data infrastructure engineers have been solving for a decade — durable state, event ordering, replayable history, observability at scale. They're building the same layer of the stack from opposite ends. With Current now part of IBM, the case for an independent, practitioner-first venue matters more than ever. Same taste. Bigger scope. CFP is open through June 30. If you're running this in production, come tell the story. streamnative.io/blog/announc… Happy Friday. Hope to see you in SF.

Six days ago I said we're a #Kafka company, too. Not an April Fools' joke. Today: introducing #Lakestream — an architecture where streams are first-class lakehouse primitives alongside tables. First proof point: a Native #ApacheKafka Service powered by #Ursa. Now in Limited Public Preview. 🧵