· Wonna, Elimäenkatu 17-19, 00510 Helsinki
An evening of talks at Wonna: iterative Bayesian modelling with PyMC to forecast Korkeasaari visits, and a live Python performance lab for LLM workflows. And pizza, drinks, and networking.
Sponsored by Wonna.
How do you turn uncertainty into decisions? In this talk, Teemu Säilynoja walks through a practical Bayesian workflow—from priors to posterior predictions—using a real-world case study: daily visitor counts at Korkeasaari (Helsinki Zoo). You’ll see how to build and iterate probabilistic models in PyMC, orchestrate them in Marimo notebooks, and reason about causality (rain, events, and even tiger-cub hype 🐯) .
Most of us don’t need “magic agents” to get real work done—we need stable, fast, and cheap LLM workflows. In this talk Allan Nevala shows practical patterns that tame hallucinations, control costs, and improve reliability, plus a live benchmark dashboard (“model kitchen”) that scores models on real-world document images.
This event has taken place. The original event listing is on Meetup.