At Breakthrough 2025, Roche’s Jorge Carmona Toscano will join a panel discussion on emerging IT strategies to maximize the value of data and bed in greater GenAI use.  

For the last 5-6, Roche has led the way in radically transforming the technology foundation that underpins its drug development operations. In other words, the core IT that supports our clinical trial activities, patient safety, the Regulatory function, and associated health authority submissions.

This ongoing transformation, toward data-driven decisions and improved competitiveness through innovation, is something I’ll be discussing in a panel session at Breakthrough 2025.

From an AI-ready backbone to pre-emptive safety 

A big part of our transformation up to now has been updating our technology backbone and standardizing our data across the organization, with a view to greater integration.

Through our collaboration with ArisGlobal, we’ve established a strong foundation for managing patient safety cases – to the point that in 2026 we expect to be at least 50% more efficient at it.

But this isn’t just about improved efficiency. It’s about enabling large-scale cross-data analysis – and ultimately predictive and prescriptive analytics – enhancing drug safety via a stronger understanding of our data and information. It’s about getting ahead of scenarios, e.g. What will we do if X happens?

Taking AI further 

Toward this next stage of our journey, we’re now looking at strategic new investment in machine learning and large language models (LLMs).

Next developments in the market will help here:

• We are on the cusp of an explosion of LLMs now, as Google’s Gemini, Grok from xAI, Meta’s Llama, and China’s DeepSeek challenge ChatGPT.

• This increased competition will further commoditize AI, improving accessibility and driving interconnectivity between LLMs.

• This will drive new use cases.

Over the next 4-5 years, AI will transform more and more manual tasks via automation (with humans in the loop for quality control) – so human intellect can be focused on more strategic tasks.

Progressing the conversation 

For my own part at Breakthrough 2025, I’m curious to hear about the ROI my peers are seeing from their own AI initiatives, and how others are ensuring high-quality, reusable data across their operations.

I’ll also be interested to discuss others’ approaches to proactive safety, and the role AI can play in helping to anticipate and mitigate safety issues before they arise.

• Jorge Carmona Toscano is Global Patient Safety Product Line Lead at Roche. He is based in Basel, Switzerland.