With ScyllaDB, a small engineering team could focus on building their product instead of battling their databases
“Just use Postgres” is the logical path for a startup figuring out product-market fit. But Sprig – an AI-powered product research platform – hit that fit much sooner than expected. And the sheer scale of the resulting data (over 1.3T events and 75 B attributes from 15B visitors) maxed out Postgres much sooner than expected.
At Monster Scale Summit 2026, Brendan Cox (Senior Staff Engineer at Sprig) walked through the full journey: why their seemingly simple platform hit database limits so fast, the various approaches they tried, and how their small engineering team worked through these challenges.
The Complexity Behind Sprig’s Surface-Level Simplicity
At the surface, Sprig’s core product seems relatively straightforward. Customers install the Sprig SDK into their mobile apps or websites to track user actions and attributes.…