When Artificial Intelligence Takes Center Stage in Advertising

After obtaining his Ph.D. in Machine Learning from the École Normale Supérieure Paris-Saclay in 2016 (Center Borelli - former CMLA), Rémi Lemonnier (Class of 2013) co-founded Scibids Technology with Julien HIRTH.

Scibids Technology is a startup that develops Artificial Intelligence whose application enhances the return on investment (ROI) of digital advertising campaigns for advertisers worldwide. Here is the portrait of an entrepreneur with a meteoric rise who retains his passion for algorithms and a desire to venture into the unknown.

Mathematics + École des Ponts ParisTech + Ph.D. from the École Normale Supérieure Paris-Saclay in machine learning: this might be the perfect equation to become a leader in online advertising purchasing. Rémi Lemonnier stumbled into this world somewhat accidentally during his Ph.D. in the mathematics department at the ENS Paris-Saclay. He seized the opportunity to pursue a thesis (CIFRE) and became an employee of Numberly (1000mercis Group), specializing in digital marketing.

The Perfect Algorithm

The thesis focused on modeling the audience for online advertisers' ads. "At the time, we could still track the browsing history of internet users. So, I modeled the targets to reach them and sell online products," recalls Rémi. "It involved solving various mathematical variables and globalizing them to optimize online advertising campaigns. The context is a bit technical," he continues. "Online ads occur in real-time through an auction system, similar to financial markets. When an advertiser has a budget to spend on advertising, they must develop a bidding strategy that takes into account a set of variables much broader than the audience.’’ Rémi found the perfect algorithm, one that combines all possible variables and allows building the best-personalized bidding strategies for any advertiser worldwide. He also published in top machine learning conferences.

Researcher and Entrepreneur: Same Battle!

With one foot in research and the other in the business during his three years at ENS, Rémi Lemonnier quickly moved towards creating his startup. "I am convinced that research and entrepreneurship have similarities. First, working independently on a long-term basis. As a researcher, you have the choice to get up every morning or not. No one above you to direct you and give you a deadline! The same goes for the entrepreneur who must work in the fog for several months, asking the same question: am I doing this for something?"

Researcher and entrepreneur also work a lot and for a long time before getting feedback. Transitioning from a system of lectures where he was taught knowledge to another where he was asked to produce it himself, Rémi Lemonnier explains, "Research is a bit like a school of self-confidence. There are always many people with more experience, resources, or specific skills on a subject. But you must believe that by focusing on a particular subject or attacking a problem from a certain angle, you will be able to make original contributions and build a stronghold in which you will become one of the best specialists. When you create a business, you must not be afraid to jump into the unknown without knowing if it will succeed."

Autonomy and Analysis: Qualities transmitted by École Normale Supérieure Paris-Saclay

For Rémi, this autonomy and freedom are success factors that École Normale Supérieure Paris-Saclay allows to develop. "ENS teaches an analytical mindset that I use every day. Normaliens consider it common sense, but by confronting different cultures, you realize that it is a distinct advantage of the French 'grandes écoles' system. Of course, we also have a lot to learn from these other cultures, such as Americans on the business sense." Among the other qualities of Normaliens, Rémi also readily cites their ability to make decisions. "It must evolve very quickly as the company grows," he notes. "The entrepreneur thinks about it a lot because it is part of the corporate culture. Should the decision be made by the loudest? Or by the most experienced? Should they be taken with a certain rigor? "We quickly formalized the fact that we make our decisions in a decentralized way and based as much as possible on data."

The Alliance of Advertising and AI

Rémi created Scibids (an alliance of Sciences and Real-Time Bidding = real-time auctions) in 2016. Developed and maintained by a team of data scientists, Scibids' Artificial Intelligence solution connects to the major programmatic media buying platforms on the market and generates precise bidding models that offer performance gains ranging from 30% to 70% for online ad buyers. It also allows the personalization and automation of digital media buying. "Custom algorithms can double the multimedia productivity of the client from one day to the next," enthuses Rémi. "Attention, our solution does not do personalized tracking," he emphasizes. "We only use the datasets of advertisers collected during each of their digital campaigns, which we then optimize to create customized buying strategies for each one." Today, companies need to personalize their advertising strategies beyond what off-the-shelf solutions on buying platforms can offer. "Our success is that we really manage to personalize AI for the goals of each online advertiser."

Operating in about ten countries, the company has experienced rapid development in recent years. It currently employs 72 collaborators. "Regardless of the advertiser's nationality, it's technically exactly the same: we offer our clients a global software solution." Who are these clients? Major brands that communicate a lot online: major brands of everyday consumer goods, appliances, or computers, as well as insurers. Scibids continues to refine its offer. "It's quite funny, in fact," confides Rémi Lemonnier. "At the very beginning, we wonder what to do once the product is developed. And then we realize that the amount of work done to improve the solution grows exponentially. Today, I don't try to anticipate our roadmap in detail beyond a few months. I know that, no matter the resources of our product team, we will always have far more important subjects to address than the capacity to address them." 

Looking for Entrepreneurial Souls

Scibids is seeking specific profiles to support its development. ‘’While the excellence of education holds significant weight, we are particularly interested in individuals with an entrepreneurial spirit. We value a profile that is not only technically sound but also embraces curiosity. Can you challenge us with questions like: Why are you asking me this? Could I approach this differently and still solve your problem? Moreover, has this production truly served its purpose? Can I improve it? In essence, we are looking for passionate individuals who consider Scibids as their own venture," asserts the CEO. 

A word to potential candidates: Rémi admits to setting the bar quite high in terms of expectations. "I prefer to wait, even for a long time, even if it's frustrating, to find the right person for the job." He often recruits from the ranks of Normaliens, with whom the connection remains strong, especially among graduates of the Center Borelli at ENS who become data scientists for the company.