Michelin Group, renowned globally for its premium tires, has long transcended its legacy as a tire manufacturer to become an industry leader in sustainable mobility, innovative materials, and technology-driven solutions. With operations spanning over 170 countries and an annual production exceeding 180 million tires, Michelin is consistently recognized for pioneering innovation.

In this third installment of a three-part series, Ambica Rajagopal, Group Chief Data and AI Officer at Michelin Group, speaks with Julian Schirmer, Co-Founder at OAO, to delve deeper into the practical lessons learned from Michelin’s ambitious AI journey. Following earlier discussions on optimizing operations, enhancing customer interactions, and building scalable data products, Rajagopal now shares valuable insights drawn from both successes and setbacks, emphasizing the critical importance of foundational readiness and strategic prioritization for ambitious AI initiatives.

Edited Excerpts

Q: Could you share some practical learnings from initiatives where you tried something, it didn’t work out as planned, but you gained valuable insights that helped reshape your AI journey?

On the AI front, we initially invested heavily in evangelizing the technology and building internal networks. Given Michelin’s DNA as a research-focused company, this resulted in numerous explorations and significant enthusiasm across the group. Many individuals actively began using generative AI for productivity. Our goal now is to elevate these individual explorations into broader, foundational projects addressing larger value-chain transformations.

For instance, we’re exploring how foundational models can significantly enhance forecasting capabilities beyond our current scale. While we’ve successfully scaled several initiatives, our ambition is now higher, focusing on larger, more impactful deployments. The idea is to push our ambition further either by extending our reach across the value chain or leveraging cutting-edge technology.

Q: Ambitious AI projects often start strong but struggle when the return on investment doesn’t immediately materialize due to organizational readiness gaps. How do you overcome these initial hurdles to realize larger AI goals?

It requires a fresh, holistic approach. Utilizing foundational models, especially for complex tasks like time-series forecasting, becomes viable only after smaller-scale experiments establish a foundation. Once the organization reaches sufficient maturity through multiple smaller deployments, there’s a shared understanding among technical and business teams, crucial for taking on larger challenges.

Additionally, these smaller initiatives clarify essential data requirements — quality, volume, and variety — forming the data foundation necessary for scaling. Thus, ambition in AI must always be accompanied by strong prioritization at the group level, ensuring we direct efforts toward the most impactful areas.

Q: How do you ensure your AI initiatives remain aligned with Michelin’s broader strategic vision? What’s your approach to prioritizing different projects?

Ensuring alignment involves two key strategies. First, we actively prevent getting locked into a narrow path based on past explorations. We achieve this by investing continuously in expertise, staying at the forefront of technology trends, and engaging extensively with academia, startups, and research institutions. This openness allows us to pursue select high-risk, high-reward projects — long-term bets on transformative technologies. For instance, we began exploring generative AI applications in tire design well before it became mainstream in late 2023.

Second, prioritization emerges naturally through smaller-scale initiatives. These projects help sensitize both technical and business stakeholders to potential benefits, risks, and the consequences of not adopting certain technologies. This shared understanding forms the foundation of our prioritization strategy.

Q: Given your active role and leadership, could you share one key piece of advice for women aspiring to careers in data and AI?

My advice, which is gender-agnostic, is to consistently invest in developing your craft and expertise. Find areas that genuinely spark your curiosity, cultivate deep technical skills, and maintain continuous learning. Technical expertise, even more than competence, is a powerful antidote to unconscious bias. Once expertise enters the conversation, all biases, including gender, tend to leave the conversation.

CDO Magazine appreciates Ambica Rajagopal for sharing her insights with our global community.