Data in Practice: Climate Intelligence and Statistical Computing with Polars

Hello PyData Milano!

On November 19, we’re meeting at TeamSystem in Milan for two practical talks about turning data into useful, dependable results. We’ll explore how AI agents can bring together weather, satellite, and hazard data to make climate risk easier to understand, then see how statistical computations can stay fast and lazy inside Polars. If you work with data pipelines, machine learning, or AI applications, join us for an evening of ideas you can take back to your projects.

📅 When: Thursday, November 19, 2026 📍 Where: TeamSystem, Milan

🪪 To access the venue, you’ll need to provide your first name, last name, and email address.

Talks

From Data & Pixels to Risk Context: Agentic Earth Observation for Explainable Climate Intelligence
Speaker: Claudio Giorgio Giancaterino
Description An MCP-enabled multi-agent Earth Observation framework combines weather, satellite, forecast, and hazard data to produce explainable, uncertainty-aware climate risk signals.

Vectorized statistical distributions, built on the Polars engine
Speaker: Francesco Bruzzesi
Description What if each row in a Polars query could have its own statistical distribution—without sending the data out to NumPy or SciPy and breaking the lazy pipeline? Francesco presents a Rust-backed Polars expression plugin that brings statistical distributions into the query itself, and shares the engineering lessons behind its performance, correctness, and reproducible sampling.

***

See you there! 🚀 The PyData Milano Team