Sébastien Burdin stands at the forefront of practical AI implementation, transforming theoretical potential into measurable business value at Swissquote. His background bridging software engineering, program management, and agile coaching positions him uniquely to navigate the complex cultural and technical challenges of enterprise AI transformation in highly regulated financial services.

The Exponential Acceleration Reality

Burdin delivers sobering statistics that cut through AI hype and force leaders to confront exponential change: « Training compute is being increased every year five times » while inference costs experience « 90% reduction » annually. Perhaps most critically for strategic planning: « The update cycle, which is accelerating. So, before we were on a new release cycle of about 6 months. More recently, it went into three months release cycle, and there are even hints that it's going to be faster than that. »

His conclusion challenges traditional strategic planning approaches: « There's no major signs of slowdown. So all of these numbers, to showcase that we're really in a new VUCA world. For those who do not know it, VUCA is Volatility, Uncertainty, Complexity, Ambiguity. And that change and disruption are happening very fast and we need to adapt. We need fast experimentation. »

The Five-Pillar Agile AI Framework

Through extensive experimentation and organizational learning, Burdin's team developed a systematic approach built on five foundational pillars. « Iterative experimentational learning is really at the core, at the base of it. Then cross-functional collaboration, really need to bring all the expertise, be it on the AI front, with our quantitative research teams, or on the business front, with whatever business expertise you have in your company. We need to bring those people together. »

The framework emphasizes risk-aware implementation: « We need to build foundation to have risk-aware agility. We need ethical framework. We have been building this to really give a frame to all our AI initiatives. We need data readiness. Data is really at the core of everything. »Finally, governance proves essential: « We really need an agile governance. We need strategic alignment, but we need sufficient empowerment of the people at the bottom up to really foster this AI transformation from the bottom. »

Practical Implementation at Scale

The Swissquote approach emphasizes systematic capability building with measurable outcomes: « We gave our participative training for more than 100 employees. We did workshop involving many people from many different departments and overall more than 200 people. And we have a very active AI community who is currently running more than 10 experimental projects. »

This community demonstrates practical innovation: « One challenge that we had is what is the best outcome we can have from a very lightweight quantized open source model running on a GeForce 1480. This was interesting because we had quite a long lead time before we could provision more capacity in our Kubernetes cluster. So this was a way to start fast and experiment fast. »

The community extends beyond technical experimentation: « We do landscape monitoring. People who have greater expertise monitor the AI news, research papers, trends forecasting, and share this in a vulgarized fashion to all the rest of the organization. And we really have a lot of internal playground experimenting around MVP, RAC projects, internal tools to enhance with AI various fields in the company, from HR processes to basically any processes that can benefit from AI. »

Change Management Philosophy

Burdin's approach emphasizes cultural transformation alongside technical implementation: « Change management starts I think from Leadership. So leadership needs to bring the space, the opportunities for change management and the possibilities for people to shift culture. » His methodology balances structure with experimental freedom through « both a combined top-down and bottom-up approach by fostering all knowledge and fostering desire from everyone across the board to participate in this new AI venture. »

Rather than mandating specific tools, SwissQuote enables experimentation: « We gave flexibility to the employees. We didn't force, for example, one single model or approach to code generation, but we enabled people to win, of course, boundaries that are imposed by the fact that we are human-regulated and so on, but to experiment with various coding generation APIs and so on. And this really was great in terms of learning experience because we realized that on some of our stack, some of those were outperforming others and vice versa in other areas. »

Vision for AI-Enabled Transformation

Burdin's vision extends beyond productivity optimization to fundamental capability transformation: « AI is going to be everywhere is going to be kind of the nervous system on steroids of all those complex adaptive systems. And I think we won't have the choice really to use it and to be part of it. »

His advice to leaders proves both inspirational and practical: « Even if you don't see clear application in your professional context, think about your dreams. Think about whatever you have dreamed of achieving, of realizing as a product or as a solution, that you don't have the competencies or the knowledge to do so. And today, the friction to be able to achieve this goal is becoming less and less of a resistance. We are having agentic systems like Flow With, I don't know, for example, if you're part of this, that within a single prompt can generate really outstanding products. So really think about your dreams and start experimenting with the AI to try to realize some of your dreams. »