Tuesday 29th September | 14:00-17:00 | Marina Bay Sands
Most teams building AI features today are stuck iterating on instinct, adjusting prompts, switching models, tweaking parameters, without a reliable way to tell whether any of it is actually working. This hands-on workshop, hosted by ClickHouse, is built to close that gap.
Led by Muhammad Ali, Senior Solution Architect, this session walks engineers through instrumenting a real TypeScript LLM agent from the ground up using Langfuse. Participants will move from having zero visibility into their AI application to a fully functioning engineering loop, complete with end-to-end trace visibility, version-managed prompts, automated monitoring for out-of-scope requests, and a reusable evaluation workflow for testing future changes.
This session is built for engineers shipping LLM features who want real answers about what's happening in production, as well as tech leads evaluating observability tooling for their AI stack. Attendees should be comfortable reading TypeScript; no prior experience with Langfuse is required.