From answering simple data questions, building cool visualisations, to building machine learning solutions and bringing them to production, these are all skills expected from data science teams.

There is so much material these days online about cool data projects that it feels like working with data is easy and projects can be finished in no time. In order to implement data products and bring them from prototype to production successfully there is though a lot of work happening behind the scenes.

This talk focuses on my personal experience of working as a product manager for a cross disciplinary data science team. I will follow a machine learning project from infancy to going live, focusing mainly on tips and tricks and lessons learned. For example, talking to stakeholders, understanding the problems they have and then translating that to actual data problems is not as trivial as it may seem. This is a talk about the challenges in tech and about the value of communication.