Like I said before that, I was mostly an individual contributor, so then I had to really, when starting a company, learn how to think long-term, how to think about roadmaps.

And in our case, the way to work is running experiments, right? So every single week you define experiments, you define different goals that your model is supposed to achieve.

The problem here is that, I guess in comparison to the typical software engineering,

you cannot force an experiment to do what you wanted to do.

So that makes planning interesting and challenging.

My name is Maria Meyer and I'm a CTO and co-founder at

Phantasma Labs based in Berlin, Germany.

This is Code Story, the podcast bringing you interviews with tech visionaries who share in the critical moments of what it takes to change in industry and build and lead.

A team that has your back.

I'm your host, Noah Laphart, and today how Maria Meyer created the tech to make data more human and model pedestrians in traffic.

All this and more on-code story.

Maria Meyer grew up bilingual as her mom was from Poland and her dad was from Germany.

In her free time, she participates in Olympic

weightlifting, which she clarified does not mean she participates in the Olympics. She finds

this is a great way to blow off steam, and outside of this, she's a bit of a foodie and

love to travel and try different restaurants. In her words, the way she started her current venture

is kind of unusual. Formerly, a back-end engineer, she got restless

and figured she was as intelligent as other founders she was observing, so she wanted to give it a shot.

Through an accelerator, she met her co-founder and built learning models to illustrate

how pedestrians act in traffic. This is the creation story of Phantasma Labs.

We are a company that heavily uses reinforcement learning, a type of machine learning,

and we build models of how human beings move around in traffic,

...