TITLE: The human side of data analysis: a tech startup perspective

Dr Flavien Lambert
Data Analyst
Piktochart

 

Abstract:

Most of the communication related to Data Science usually refers to well-defined problems and limited scopes, e.g. how to choose the best models and fine tune the parameters. Few though explain that Data Science (or Analysis) practitioners do not work isolated but are embedded into business or marketing (even design!) teams whose backgrounds lay far from machine learning or statistics in general.
In this talk, I will highlight the specific environment of a small tech startup and how it shapes the day-to-day activity of those in charge of analysis. I will focus, in particular, on three main aspects:

• the interaction with the developers teams to shape the collection of data;
• the challenges in exchanging with non-technical teammates on both defining the problem and answering it;
• the unsettled place of Data Science/Analysis in a company whose product does not rely on Data Science per se.

By the end of this presentation, you will hopefully understand that a small company can be a motivating space for scientists, space which calls for a very broad range of (soft) skills and knowledge beyond algorithms.
 

Biography

Flavien spent the first eight years of his professional path at the French Atomic Energy commission where is studied the modelling - at the atomic level - of matter in extreme conditions and helped in designing LASER experiments.  After a year exploring South East Asia, he joined, in 2004, a team from the MIT in Singapore to tackle problems related to urban planning. To this aim, he made use of cellphone records and other digital traces to understand mobility patterns in the city-state. Since August 2015, Flavien is in charge of the Analysis team in Piktochart. He helps his co-workers to get accurate information in topics as different as finance, customer behaviour or design.