
Fabrice Bernhard
Founder of Theodo, a tech consultancy he scaled with Benoît Charles-Lavauzelle from 10 to 700 people in 10 years.
He worked on his first large legacy modernisation in 2008 and has been back in the trenches to experiment first-hand how AI is revolutionising this challenge.
Fabrice has shared his Lean Tech and AI modernisation expertise at various conferences, including CraftConf, DevopsDays and Lean Summits.
Lean Tech: How to Lead on Creating More Value With AI
Lean Tech: How to Lead on Creating More Value With AI

The Hype vs. RealityTrillions of dollars invested in AI are fueling massive excitement. Yet amid the buzz, it's tough to separate fact from fiction. Science paints a sobering picture: MIT reports 95% of AI investments deliver no value, while METR finds AI slashing productivity by -20%. Sound familiar? In your organization, vocal AI advocates push big claims, often sidelining skeptical senior engineers.The Problem in Your TeamsThis dynamic risks wasting resources and alienating expertise. How do you cut through the noise and pinpoint where AI truly creates value?Lean Tech: The Scientific AntidoteEnter Lean Tech—an adaptation of Lean Thinking for tech, rooted in Toyota's methods and Deming's scientific approach to work. It's the ideal framework for navigating AI's revolution, focusing ruthlessly on value creation.Key Lean Tech Principles in Action (from Theodo's Experience): - Value for the Customer: Spot AI-solvable problems and measure improvements rigorously.- Tech-Enabled Network of Teams: Empower autonomous teams to experiment with AI tools.- Right-First-Time: Analyze every issue, no matter how small, to extract lessons.- Just-in-Time: Track lead-time gains for productivity, not just cycle time.
Building a Learning Organization: Use 6-step Kaizens, standards, skills matrices, and dojos to scale knowledge.Proven Outcomes: 3x faster legacy modernizations and 2x acceleration on projects where product decisions aren't the bottleneck.Attendees will walk away with a step-by-step playbook to inject science into AI adoption: identify high-impact opportunities, measure real ROI, foster autonomous experimentation, and build lasting learning systems—ensuring your teams deliver tangible wins without the hype.
Building a Learning Organization: Use 6-step Kaizens, standards, skills matrices, and dojos to scale knowledge.Proven Outcomes: 3x faster legacy modernizations and 2x acceleration on projects where product decisions aren't the bottleneck.Attendees will walk away with a step-by-step playbook to inject science into AI adoption: identify high-impact opportunities, measure real ROI, foster autonomous experimentation, and build lasting learning systems—ensuring your teams deliver tangible wins without the hype.
AI-accelerated Legacy Modernisation
AI-accelerated Legacy Modernisation

Our world runs on IT built in the 80s and 90s by engineers who are now retiring. This is becoming very worrying for corporates who have critical systems running on legacy code that only a handful of people still understand.That's where GenAI arrives. With its almost magical transpilation abilities, it is the innovation that was needed to kickstart the legacy modernisations that had been postponed for too long.Having worked on critical modernisations for more than 15 years, I jumped on the opportunity to experiment with how LLMs can accelerate complex migration projects. I will share the concrete experience we have accumulated on a wide range of stacks (Swift->React-Native, Eclipse RCP -> Spring Boot, Java 1.6 Spring -> Java 21 Spring Boot, PHP ZF1 -> Symfony 7), cover both the good surprises and the limitations that we uncovered, and share our current playbook to best leverage AI when migrating a legacy system.