First presentation

Things we've learned working with a lot of Timeserieses

A Local-First approach with great MLOps, visualizations and over all tooling.
We'll share some of our best tips through examples where we've failed and then proven ourself, and sometimes where we did a smash-hit from the get-go!
The talk will include things like Local-First approach, thinking about efficiency, MLOps, interactive visualization tooling and how you should think about building your own loss-functions if your use-case is unique.

Second presentation

Surface classification with radar data

Radars are often associated with detecting objects such as approaching aircrafts, or measuring the speed of moving vehicles. But there are plenty of less known radar applications that have emerged alongside advancements in machine learning. In this presentation we will see how radar measurements and machine learning were used by an autonomous lawmower to recognize grass surfaces.