Moving From Hype to Solutions

We recently returned from BioTechX Europe 2025, and the experience confirmed a significant shift in the pharmaceutical landscape. Unlike other events we attended this year, BioTechX proved to be a deeply technical gathering centered on the hard scientific and engineering challenges of modern drug discovery.

The conversations have moved beyond the theoretical "what if" of AI to the practical "how"—specifically, how to integrate these powerful tools into established enterprise workflows. Our team was on the ground contributing to two major tracks: the implementation of FAIR data principles and the realistic evaluation of Large Language Models (LLMs).

FAIR AI for Evidence Synthesis

Our CTO, Artur Nowak, took the stage alongside Seye Abogunrin, Global Access Evidence Lead at a major pharmaceutical enterprise, to present a new workflow for tackling the data interoperability crisis.

They demonstrated how "Agentic AI" combined with human-in-the-loop governance can ensure data is "born FAIR"—structured and standardized from the very moment of extraction.

[Read the full summary of the presentation and the "Born FAIR" workflow here.]

Panel: Are LLMs Enough?

In addition to the presentation, Artur joined a panel discussion titled “Are LLMs enough?” alongside leaders from Tangram Therapeutics, Bayer, Roche, and SAS.

The group explored the role of LLMs not as standalone magic wands, but as powerful components within carefully designed systems, debating the necessity of human baselines and the potential for AI to improve compliance.

[Read our key takeaways and reflections from the panel discussion here.]

Final Thoughts

The overarching theme of BioTechX 2025 was that the industry possesses strong in-house technical capabilities and is now seeking enterprise-grade integration rather than simple point solutions. We are excited to continue building the infrastructure that allows these technical teams to turn unstructured information into strategic, interoperable enterprise assets.