Path to power
My studies were driving me to research in formal logics and type systems, when I fell into start-up world in 2000. I was 20 back then, and I started working at a search engine company, where I was managing engineering. Neuro-linguistic programming was for me a great way to learn about data architecture and machine learning at scale. I then worked on building data science teams in various tech companies before founding Dataiku in 2013.
What is the proudest achievement of your career to date?
My proudest achievement has been watching Dataiku grow from a small start-up into what it is today – one of the leading AI platforms in the world, with more than 300 customers and 400 global employees. Seeing how Dataiku can change businesses and help them incorporate AI into their processes and products is a close second. I believe that AI is the most significant business transformation opportunity since the invention of electricity and being a small part of that with Dataiku has been extraordinary.
Who is your role model or the person you look to for inspiration?
I don’t really have a role model as one given person I would get inspiration from. I read books, fiction and non-fiction, without expecting them to be all real. Inspiration comes from good stories, I think.
Did 2019 turn out the way you expected? If not, in what ways was it different?
In many ways, 2019 did turn out the way I expected. We saw a huge focus on responsible AI, including explainability of AI decisions, white box models, ethics, etc, which is something that we at Dataiku have been working on for quite some time and have expected would come to the forefront.
What do you expect 2020 to be like for the data and analytics industry?
I think that 2020 will be the year for mature cloud usage for analytics in large enterprise. Cloud has been pervasive already in tech companies, but its widespread adoption for large enterprise was lagging behind for analytics, given the cost of migrating data from legacy systems and the security constraints associated to sensitive data. We’re already starting to see the shift, but this year is where it will really take off. The most advanced organisations will take a hybrid approach, mixing multiple cloud and on-premise solution for best cost and security trade-offs.
Data and technology are changing business, the economy and society – what do you see as the biggest opportunity emerging from this?
Most forget that at the centre of all this change, advancements in data and technology, AI, etc, there are still people. So, I think for me, the biggest opportunity is for organisations to use AI not to replace humans, but to augment them. AI is still no match, in many ways, for people, and the power to use humans and machines together is far greater than the opportunity for machines alone.
What is the biggest tech challenge your clients face in ensuring data is at the heart of their digital transformation strategy?
Sometimes just making the very first step, having data and making good use of it is the biggest challenge. But taking a forward-looking glance, scaling AI is the biggest challenge clients face today. That is, going from pushing one or two models in production a year to thousands. It’s hard on many levels, one of which is simply being agile enough from a personal perspective. Technically, there’s a need for consistency from the design of AI systems themselves, so that organisations can get to the next level: automatise the maintenance of AI itself.