Pavan Keshavamurthy

The year is 2026. Self managed still owns in the controlling stake in the Kafka market. Banking and Financial Services, Gov and other industries continue to self-host, particularly under regulation guardrails. Long-tail cloud infrastructure providers and on-prem is still the cheapest price tier per GB/s of data streamed, often by orders of magnitude relative to hyperscalers In this context arises the need for a hybrid management stack to run your private streaming estate, whether it is on BYOC or just plain Kubernetes This session will cover Streamtime and our experience of building out an enterprise grade private cloud. Streamtime is a meta-harness that integrates perations for Kafka holistically through a management plane above OSS Kubernetes operators. - Install: One management API contract to rule it all: Spawn and delete clusters anywhere. Provider-aware (both OSS and commercial distributions) and infrastructure aware - Support REST API, CLI, Terraform -- all through one contract - Operate: Upgrade versions, roll configuration changes, and more - Troubleshoot: Reason deeply (with native integrations to AI) about fleet/infrastructure, kafka metrics and handle incidents without…

This session is designed to make Apache Flink accessible to anyone eager to explore streaming and batch processing—without the barriers of complex setups or high costs.