Build a first machine learning component with a product-first approach by Raphaël Korach
1 min
by Christophe Noujaim

Raphaël Korach, VP Product @Getaround, explains how launching a new machine learning component for smart pricing starts with answering a design and marketing problem. He introduces the Four O framework the team used to strike the right balance between automation and manual actions, then proceeds to explain why starting narrow and fully featured is the way to go in machine learning.
The main key learnings :
- Focus on design on Marketing. Implementing a new machine learning component goes beyond technical subjects
- Learn the Four O framework. Balance automation and manual actions with the Four O: Organisation, Oracle, Observation and Optimisation
- Start narrow instead of a large perimeter. In the case of machine learning, you should start narrow and fully featured instead of large and less featured
To learn more about Product Management: download our book Agile Product Management
Written by

Christophe Noujaim
Christophe started his career as a Product Marketing Manager at Biosanté. He continued as a product manager at Gandee. After a brief experience as a data account manager with Effilab, he returned to product management at Comptalib and Bloomflow. He became a product ops at Accor, where he supported the implementation and optimization of product processes for a year and a half.
Lets stay in touch, sign up for our Product Newsletter
Stories, analysis and feedback to help you grow as a Product. Once a month, no more.