Keewano DB, a Database Built for Machine Reasoning at Scale
A database for machine reasoning is a very different animal from a warehouse built for dashboards. Mark Kardashov, CEO and co-founder of Keewano, joins Alan Shimel on Techstrong TV. Furthermore, he explains why agents needed a storage format designed from scratch.
About Mark Kardashov
Mark started 20 years ago as a software developer and led a small DBA team on Microsoft SQL. In addition, Keewano is his third company after Test Project, sold to Tricentis, and a 300 person integration firm. Consequently, he returned to his database roots at exactly the right moment.
Why a database for machine reasoning
The idea came from the CTO and general manager at Plarium, a large gaming company. Meanwhile, they argued that machines were becoming the biggest consumers of data. Therefore, a database for machine reasoning had to replace middleware bolted onto old infrastructure.
Mark saw an early engine written in pure C running on a laptop and was sold. Furthermore, the team rebuilt everything from storage to engine to interfaces. As a result, Keewano DB answers queries that were never optimized for in advance.
Who Keewano DB is for
Keewano targets analytical workloads rather than transactional databases. In addition, it fits fintech, e commerce and any B2C product with billions of events. Consequently, teams reason on live data instead of aggregates that arrive days later.
Mark also sees a fit for behavioral observability of agents at scale. Meanwhile, log and textual observability is planned for a future release. Therefore, teams can trace which tools were called and what led an agent astray.
Availability and pricing
Keewano DB reached general availability last week after two years with design partners. Furthermore, pricing charges per entity rather than per event, and a plain PostgreSQL interface works too.
Explore more artificial intelligence coverage and the latest Techstrong TV interviews.
For more information please visit keewano.com
