Meet Loic, Datavillage's Principal Data Scientist

Submark datavillage logo
Datavillage Aug 06, 2021
Loic Datavillage Data Scientist

Can you please introduce yourself and tell us more about your background ?

My name is Loïc Quertenmont and I have a doctorate in high energy physics. I started my career as a research physicist analyzing the 50Pb of data produced yearly by the CERN LHC in view of finding evidence of new physics. Among other things I had the privilege to be involved in the discovery of a new particle: the Higgs Boson. I am now exploiting my 15 years of expertise in data analysis and software development to help companies maximize the insights of their data. So far, I was involved in projects related to customer analytics, geo-spatial, food, retail and utility sectors.

What is your role at Datavillage ?

My role, as principal data scientist at Datavillage, is to develop processes and algorithms that combine the vision of datavillage for data privacy and control with the business needs of the clients and prospects. Adding data privacy and control into the equation often requires rethinking how typical machine learning algorithms are built to avoid accessing or leaking user data. Another important part of the role consists in the enrichment of the user's personal data with external (open) data sources which help to get the full picture. This is extremely exciting because datavillage is really a pioneer on all these aspects !

Give us your predictions, what do you think technology will be in the next 10 years and what are the challenges of today ?

I believe that the decentralization of data will carry on and will soon be joined by the decentralization of computing power. Today, the majority of the computing resources are already in the cloud but are still owned by only a small group of actors (GAFA and more). I expect that more specialized actors come into play and that interaction between all these actors will be unavoidable. The major difficulties I see with this evolution is about data sharing, standardisation, persistence, privacy, permissions, etc. We need to think today how to technically gear up all these actors together and more importantly how to let the user decide for himself how his data could be used and enriched, for what purpose and by whom.
From the data science point of view, one of the main difficult with this paradigm shift is that no actor will ever have access to the full data picture of a user, so we need to figure out the best strategy to “enrich” the user data with external sources in order to recover and even extend AI capabilities of today’s algorithm while preserving data privacy.

Why have you chosen to work with Datavillage ?

Datavillage is trying to answer these complex questions. They are not the only ones, but they have a pragmatic approach that improves user data privacy and most importantly as of today they are already creating value to both their users and their customers (service providers) thanks to this shift of paradigm. I like this strategy and I’m glad to be part of the adventure.

According to you, what is the number one challenge of digitized organizations today ?

The main concern today is that they are two kinds of “digital” companies. The GAFA’s and other major digital actors are massively centralizing user data and do all their best to keep them and restrict access to those data to other actors. Their main business is the data itself and they are offering other services mostly to accumulate more and more data. The other companies are experts in their domain and most of the time have very little knowledge and experience about data utilization. They realize that data and more generally digitization are necessary to provide better services to their customers but they don’t really know how and where to start the journey. It is even more difficult as the major data companies (and sometimes regulation) make their road even more complex.
The situation is a bit similar to what happened a decade ago between closed source software (mostly owned by a few leading companies) and the open source software community. Nowadays, an open source equivalent of almost all tools is available and many people are relying on it. Even the major software companies release most of their tools freely. The open data is on its way but still has a lot of open questions regarding “personal” data. I believe the “open algorithm” or “open processing” will come soon and help small actors of the digital world to compete with today’s majors.

Submark datavillage logo