As an architect, I’ve seen brilliant systems fail under the weight of poor data-layer choices — and witnessed unexpected success from surprisingly “wrong” decisions. With over 50 years of database evolution behind us and hundreds of engines listed on db-engines.com, it’s time to reflect.

This talk is a deep dive into the evolution of database systems — from relational beginnings and NoSQL rebellion to NewSQL, lakehouses, vector stores, and embedded edge databases. We’ll unpack the architectural ideas that changed everything, from ACID to LSM trees, from masterless to multi-model.

You’ll hear lessons learned (sometimes painfully) from real-world projects, explore ideas that changed everything, and discover some strange and wonderful systems you’ve probably never heard of.

The talk ends with a cheat sheet to help you make better database decisions — and avoid repeating history’s mistakes.