PlanetCassandra #13 - CQLAlchemy – A Pythonic Toolkit for Apache Cassandra 6.0

Stop Waking Your Senior Engineers at 3am: AI-Assisted Cassandra (and Kafka) Operations

Running Apache Cassandra at scale still leans on a handful of senior engineers who can read the tea leaves — gossip, compaction, repair, GC, replication. When an alert fires at 3am, investigation often takes 30–60 minutes before anyone even understands the cause, and that operational knowledge tends to live in two or three people's heads.

In this session, Johnny Miller looks at how AI is starting to change day-2 Cassandra operations — not as a generic chatbot bolted onto a dashboard, but grounded in your actual cluster topology, config and workload. Through a live demo, he'll walk through asking natural-language questions about a real cluster, automated root-cause analysis when an alert fires, and recommendations that surface problems before they become incidents. He'll be candid about what works today, what's still on the roadmap, and why keeping a human in the loop matters.

Johnny will be joined by special guest Patrick McFadin for the discussion — bringing two decades of hard-won Cassandra operational wisdom to the conversation.

About the speakers

Johnny Miller has been running Apache Cassandra in anger since v0.6 and has the scars to prove it. He's Co-Founder of AxonOps and Digitalis.io, where he wrangles Cassandra and Kafka at scale and all the weirdness that brings — so you don't have to.

Patrick McFadin is one of the most recognizable voices in the Apache Cassandra community. An Apache Cassandra committer and PMC member, ASF Member, and co-author of the O'Reilly book "Managing Cloud Native Data on Kubernetes," Patrick has spent over two decades building scale infrastructure and helping organizations run some of the largest Cassandra deployments in production. He's spent his career traveling the world educating developers on distributed databases.