In enterprise, you might get a stakeholder, who wants to know exactly what the system was reasoning, ask why this user got an incorrect response to the question "show me data" or simply have full statistics into what was queried, which people do it, and provide a globally approved list of prompts, that they can edit. We were able to achieve it with langfuse, using our fastapi ai engines and full-stack nextjs platform
Event details
AI Builders and Databases Krakow starts at 18:30 on 4 November 2026 at Hevre in Kraków. The time is in Europe/Warsaw.
Technologies covered
The talk covers Langfuse observability and prompt management integrated with FastAPI AI engines and a full-stack Next.js platform, including investigation of incorrect responses and statistics on queries and the people making them.