There is a conflict of interest: data science field has evolved from academic routines and is often based on long-term cycles of research and seeking for the optimal solution. On the other hand, product development requires fast execution, flexible decision making and constant deliverables of new solutions to production.

In this talk I’ll share how agile product development principles work in harmony with machine learning development on real examples from my work at Booking.com. I will explain how embedded data scientists in cross-functional team structure and data-driven culture can contribute to daily standups and constant deliverables, and how to tackle big machine learning projects in short cycles.