Just announced at Interrupt! SmithDB. Agent traces have outgrown the databases built to hold them. That’s why we built SmithDB, a purpose-built distributed database for agent observability. Read the announcement from Co-Founder Ankush Gola →

@ankush_gola11 Game changer this made me decide no Mastra I will use Langchain since seems Mastra copies Langchain & Langchain just works in many diff scenarios Deepagents Langraph & now SmithDB

@ankush_gola11 What options did you evaluate before building this?

@ankush_gola11 Could just used @helixdb … for free

@ankush_gola11 Purpose-built DB for agent traces with tens of thousands of spans and large payloads makes sense. Traditional stores choke on this shape of data. Curious about the storage format and query patterns that gave the biggest gains

@ankush_gola11 the agents are officially talking too much for postgres to handle

@ankush_gola11 huh, neat

@ankush_gola11 Purpose built trace stores are the right move. Splitting by attribute cardinality early matters more than vendors admit. The risk is a schema so narrow that next quarter's agent feature breaks every query. Curious how SmithDB handles open fields.

@ankush_gola11 LOVE!

@ankush_gola11 Traces outgrowing generic DBs makes sense. The next layer I keep wanting is the human-facing receipt on top of the trace: what changed, what passed, what failed, what is still an assumption, and what the reviewer should do next.

@ankush_gola11 the fastest agent observability database!!! 🚀🚀🚀

@ankush_gola11 the honest answer is that our db was a hack until smithdb came along. we thought we could make do with existing tech, but it didn't scale

@ankush_gola11 Who makes these videos