The Rise of the AI Database Administrator
The AI database administrator is quickly becoming a serious platform engineering topic. Itamar Syn-Hershko, CEO and founder of Neverblink, joins Alan Shimel on Techstrong TV. Furthermore, he shares why the hot serving layer needs a specialist that never blinks.
About Itamar Syn-Hershko
Itamar has spent two decades building and consulting on databases, data lakes and data warehouses. In addition, he bootstrapped a consulting practice and a product company in parallel. Consequently, Neverblink reflects lessons from years of production data platform work.
Why an AI database administrator matters
Itamar explains that Neverblink began as an automatic DBA long before modern generative AI. Meanwhile, the recent AI wave supercharged the roadmap without changing the mission. Therefore, the goal is still a system that watches, tunes and fixes the serving layer.
He notes that human DBA counts keep shrinking as expectations keep rising. Furthermore, product engineers now ship code, run production and touch the database. As a result, teams need an AI database administrator that scales across a fleet.
ClickHouse joins OpenSearch and Elasticsearch
Itamar walks through the new Neverblink support for ClickHouse alongside OpenSearch and Elasticsearch. In addition, ClickHouse now powers dashboards with millions of writes per second at real scale. Consequently, an AI database administrator has to reason about OLTP and OLAP together.
He warns that general chatbots return wrong or incomplete answers for niche systems like ClickHouse. Meanwhile, coding agents change schemas without knowing the runtime cost. Therefore, deep specialist tooling has to guard the database when the agents cannot.
What is next for the AI database administrator
Itamar and Alan close with a preview of the Cloud Native Now AI stack event in late October. Furthermore, Itamar will keynote just before KubeCon with a new Neverblink announcement. Consequently, the AI database administrator story keeps expanding across OLTP and OLAP systems.
He points to the ngrams trap in OpenSearch as a classic example of hidden database cost. In addition, this is exactly the kind of choice a specialist AI database administrator should catch. Therefore, teams gain speed without paying a silent performance tax later.
Explore more artificial intelligence coverage and the latest Techstrong TV interviews.
For more information please visit neverblink.ai
