I Hated AI… Until This Changed Everything (Real-World MCP Use at ClickHouse)
Dmitry Pavlov
What happens when a hardcore data infrastructure guy goes from hating AI to building production-grade AI agents?
In this talk, Dmitry Pavlov from ClickHouse shares how Model Context Protocol (MCP) completely shifted his view—and how he used it to give every team in his company instant access to data insights without BI dashboards.
Key Takeaways:
✅ Why most AI tools fail inside enterprises
✅ The real challenge with getting business teams to use AI
✅ How ClickHouse built their own internal AI agent (“Dwayne”)
✅ What MCP is, and why MCP 2.0 could change the game
✅ Live examples: AI querying real databases on the fly
Recorded at the Push to Main Baby! Meetup in Amsterdam.
Scaling ClickHouse to petabytes of logs at OpenAI
How ClickHouse helps Anthropic scale observability
How Capital One Slingshot cut infrastructure costs by 50%
Engineering leaders at Capital One Slingshot share how they cut infrastructure costs by 50% and reduced average dashboard load time from 5+ to under 500ms with ClickHouse Cloud.
