In this Tech & Drugs episode, I sit down with Thomas Steger-Hartmann, former investigational toxicology lead at Bayer and industry lead of the VICTOR Consortium, to explore the future of drug safety.We discuss how toxicology is moving from descriptive animal studies toward more predictive, data-driven, and less animal-dependent approaches — without losing sight of the ultimate goal: protecting patients.Thomas explains why animal models are often more useful than public debates suggest, where they fall short, and how new approach methodologies, virtual control groups, organ-on-chip systems, historical control data, and AI could reshape preclinical safety assessment.A major theme of the conversation is the VICTOR Consortium, which aims to use historical control data, statistics, and artificial intelligence to build virtual control groups and potentially reduce animal use in toxicology studies.We also discuss the role of regulators, why regulatory acceptance is often misunderstood, and how collaboration between industry, EMA, FDA, and scientific consortia can help move new methods into practice.Key themes discussed:Why toxicology is central to drug discovery and developmentWhat animal studies do well — and where they fall shortThe limits of translating animal findings to humansWhy rare adverse events are statistically difficult to detectHow virtual control groups workThe VICTOR Consortium and historical control dataHow AI can help extract value from toxicology datasetsNAMs, in vitro systems, in silico tools, and organ-on-chip modelsRegulatory science, qualification, validation, and acceptanceWhy better data curation is essential for safer and more ethical R&DWhy this matters:Drug safety sits at the intersection of biology, data, regulation, and ethics. As pharma and biotech move toward AI-enabled R&D, toxicology is becoming a critical test case for how the industry can use historical data, computational methods, and new experimental systems responsibly — reducing animal use while maintaining scientific and regulatory confidence.Guest information:Guest: Thomas Steger-HartmannRole: Former investigational toxicology lead at Bayer; industry lead of the VICTOR ConsortiumCompany: Bayer / VICTOR ConsortiumLinkedIn: https://www.linkedin.com/in/thomas-steger-hartmann-b7b05a55/Company website: https://www.bayer.com/ & https://www.vict3r.eu/Tech & Drugs explores how data, AI, and technology are changing pharma, biotech, and drug R&D. Hosted by Thibault Geoui, the podcast brings together leaders, scientists, technologists, and builders working at the interface of science and technology.If you enjoyed this conversation, subscribe to Tech & Drugs for more discussions on AI, data, and the future of pharma and biotech.#Toxicology #DrugSafety #AIinPharma #DrugDiscovery #NAMs

In this Tech & Drugs episode, I sit down with Patrick Ansems, globally responsible for life sciences strategy at Siemens across pharma and medical devices, to explore how AI, data, automation, and platforms are reshaping drug development.Recorded live in London at the Pistoia Alliance meeting 2026, this conversation looks at a central question for the industry: can pharma make drug development learn faster than it forgets?Patrick has worked across science, software, R&D, lab informatics, and manufacturing, with experience at PerkinElmer Informatics, Tetrascience, Dotmatics, and now Siemens. In this episode, we discuss why AI is forcing leaders to think differently, why FAIR data has new urgency, and why pharma still depends so much on Excel, PowerPoint, documents, and institutional memory.We also explore the “messy middle” of tech transfer, the role of context in scientific data, and the idea of “pharma in the loop” — connecting discovery, development, and manufacturing into a more integrated learning system from molecule to market.Key themes discussed:- Why AI is changing leadership in pharma and life sciences- FAIR data, AI-ready data, and why data context matters- The gap between scientific progress and slow decision-making- Why pharma workflows still rely on Excel, PowerPoint, Word, and institutional memory- Tech transfer as the messy middle between R&D and manufacturing- How to reduce corporate amnesia in scientific organizations- The role of Dotmatics within Siemens- Scientific intelligence platforms and the future of connected lab data- Lab automation, manufacturing, simulation, and AI-enabled experimentation- Why AI is not a silver bullet for messy infrastructure- “Pharma in the loop” from molecule to marketWhy this matters:AI will only create real impact in pharma and biotech if it is connected to high-quality data, scientific context, robust workflows, and the realities of development and manufacturing. For R&D, digital, data, technology, and manufacturing leaders, the challenge is no longer just adopting AI tools. It is building systems that can learn across the full drug development lifecycle.

Guest information:Guest: Patrick Ansems

Role: Global Head of Life Sciences at Siemens Digital Industries SoftwareCompany:

SiemensLinkedIn: https://www.linkedin.com/in/patrickansems/Company website: https://www.siemens.com/en-us/company/about/businesses/digital-industries/About the podcast:Tech & Drugs explores how data, AI, and technology are changing pharma, biotech, and drug R&D. Hosted by Thibault Geoui, the podcast brings together leaders, scientists, technologists, and builders working at the interface of science and technology.If you enjoyed this conversation, subscribe to Tech & Drugs for more discussions on AI, data, and the future of pharma and biotech.#AIinPharma #DrugDevelopment #LifeSciences #PharmaR&D #TechAndDrugs

In this Tech & Drugs episode, I sit down with Laura Matz, Chief Science and Technology Officer at Merck KGaA Darmstadt, Germany, to explore how AI, data, automation, and digital technologies are reshaping science at scale.Laura brings a rare perspective across chemistry, semiconductors, life science, healthcare, and advanced materials. We discuss what pharma can learn from the semiconductor industry, why AI is changing how scientists design experiments, and what it takes to move from pilots to real impact inside a large organization.The conversation goes beyond generic “AI in pharma” claims. Laura shares a practical view of how AI can help scientists make better experimental decisions, why data governance and infrastructure matter, how autonomous labs are being built, and why leadership in the AI era requires both speed and responsibility.We also explore the future of foundation models for chemistry, biology, and physics, the role of Europe in global science and technology, and what young scientists should do to prepare for careers at the intersection of science and technology.Key themes discussed:- How AI is changing the role of science and technology leadership- What pharma can learn from Moore’s law and the semiconductor ecosystem- Bayesian optimization and smarter experimental design- Human intuition vs machine-guided discovery- Scaling AI beyond successful pilots- AI-ready data, governance, and secure access- Autonomous labs and the connection between physical and digital science- How leaders can balance speed, stability, and experimentation- Europe’s role in global science and technology competitiveness- Foundation models for chemistry, biology, and physics- Career advice for scientists entering an AI-enabled worldWhy this matters:AI will not transform pharma, biotech, or R&D through models alone. The real challenge is connecting data, infrastructure, scientific expertise, leadership, and operating models in a way that helps scientists move faster while preserving rigor, safety, and trust.Chapters:00:00 Introduction01:10 Laura Matz’s background and early scientific curiosity03:25 Basketball, teamwork, and leadership05:05 From pre-med to chemistry08:00 What a Chief Science and Technology Officer does10:00 How the CSTO role changed after ChatGPT11:25 What pharma can learn from semiconductors14:35 Is pharma truly more complex than other industries?16:20 Biology, engineering, and the tension between tech and life sciences18:40 Where AI is already impacting R&D workflows19:15 Bayesian optimization and better experiment design21:15 Pairing AI experts with scientific experts22:25 Human intuition vs machine-driven discovery25:20 Scaling AI pilots beyond the “messy middle”28:15 Adoption friction in large organizations29:45 AI-ready data and the foundations of AI transformation35:50 Data access, governance, and security37:25 Building an autonomous chemistry lab in Boston39:25 Decision-making in the AI era40:15 Why leaders need to experiment with AI themselves41:10 Balancing speed and stability in AI transformation47:30 AI as augmentation, not replacement50:00 Europe, innovation, and global competitiveness52:00 Foundation models for chemistry, biology, and physics52:45 Career advice for young scientists54:00 Closing thoughtsGuest information:Guest: Laura MatzRole: Chief Science and Technology OfficerCompany: Merck KGaA Darmstadt, GermanyLinkedIn: https://www.linkedin.com/in/laura-m-matz/Company website: https://www.merckgroup.com/enTech & Drugs explores how data, AI, and technology are changing pharma, biotech, and drug R&D. Hosted by Thibault Geoui, the podcast brings together leaders, scientists, technologists, and builders working at the interface of science and technology.If you enjoyed this conversation, subscribe to Tech & Drugs for more discussions on AI, data, and the future of pharma and biotech.#AIinPharma #DrugDiscovery #AutonomousLabs #Biotech #TechAndDrugs

In this Tech & Drugs episode, I sit down with Walid Kamoun, VP and Global Head of Oncology R&D at Servier, to explore how AI is changing oncology drug discovery and development.We discuss where AI is already useful today, what remains difficult, and how pharma leaders can think about AI beyond hype, pilots, and generic “transformation” language.Walid shares a grounded view from the front lines of oncology R&D: how AI can support asset leaders, clinical scientists, target discovery, molecule design, trial planning, regulatory work, and patient matching. We also discuss why AI adoption is not only a technology question, but an operating model, culture, data, and leadership question.A central theme of the conversation is AI as a booster for expert work. Rather than replacing scientific and clinical judgment, AI may help teams create stronger “draft zero” development plans, accelerate decision-making, and focus human expertise where it matters most.Key themes discussed:- How AI is changing oncology drug discovery and development- Why AI should support, not replace, expert judgment- The role of AI in asset leadership and integrated development plans- “Draft zero” thinking for oncology programs- AI in synthetic chemistry and synthetic biology- AI use cases in clinical development, protocol writing, and regulatory work- Matching the right patient to the right drug in precision oncology- Why AI-ready data and infrastructure matter- How biotechs may benefit from AI as an accelerator- How pharma can evaluate AI partners beyond marketing claims- The role of big tech in pharma and biotech R&D- Why oncology R&D still needs strong human, scientific, and clinical leadershipWhy this matters:AI is already influencing how pharma and biotech teams discover, develop, and evaluate new medicines. But the real opportunity is not simply using more tools. It is understanding where AI can improve R&D decisions, accelerate timelines, strengthen development strategies, and ultimately help bring better therapies to patients.Guest information:Guest: Walid KamounRole: VP and Global Head of Oncology R&DCompany: ServierLinkedIn: https://www.linkedin.com/in/walid-kamoun-10223288/Company website: https://servier.com/en/servier/Tech & Drugs explores how data, AI, and technology are changing pharma, biotech, and drug R&D. Hosted by Thibault Geoui, the podcast brings together leaders, scientists, technologists, and builders working at the interface of science and technology.If you enjoyed this conversation, subscribe to Tech & Drugs for more discussions on AI, data, and the future of pharma and biotech.#AIinPharma #Oncology #DrugDiscovery #Biotech #PharmaRND

🎙 Tech and Drugs – Season 02, Episode 03From Code to Cells: Dov Gertz of Converge Bio on Generative AI and the Future of Drug DiscoveryIn this episode, I sat down with Dov Gertz, CEO and co-founder of Converge Bio, a company pushing generative AI directly into the language of biology.Dov represents a new breed of scientist.Trained in computer science and bioinformatics, with research roots in CRISPR discovery alongside leading pioneers like Jennifer Doudna, he is now building AI systems that don’t just analyze biology but actively design it.This conversation sits at the heart of what’s changing in our industry right now - From data to models to real-world impact.What we cover:✔️ Dov’s journey from computer science to CRISPR research and AI-driven biology✔️ Why the biggest bottleneck in AI for drug discovery is not compute, but data quality✔️ The shift from traditional AI to generative AI and why it changes everything✔️ How Converge Bio trains models directly on DNA, RNA, and protein sequences✔️ Why biology is fundamentally harder than NLP despite having more raw data✔️ The rise of autonomous labs and why they are critical to unlock AI’s full potential✔️ Why most clinical failures are not about targets, but about molecules✔️ How generative models can design better antibodies in a single iteration✔️ The reality of AI agents in science and why we are still far from “AI researchers”✔️ Why small, highly skilled teams are outperforming large R&D organizations✔️ The transition from AI experimentation to industrialization in pharma✔️ When we will actually see AI impact FDA approvals (and why it will take time)One idea that stood out for me: We are finally moving from AI as a tool… to AI as a generator of biology.That’s a very different paradigm.Dov brings a clear and grounded perspective on where AI truly works today, where it still struggles, and why the next breakthroughs will come from combining better data, better models, and tighter integration with experimental systems.If you care about the future of drug discovery, the role of generative AI in biology, and what it takes to move from promise to real impact, this episode is for you.I hope you enjoy this conversation as much as I did.

🎙 Tech and Drugs – Season 02, Episode 02
From Tennis Courts to Molecular Design: Tim Hoctor on Data, Discovery, and the Future of Pharma

Last December in Berlin, I had the privilege of sitting down with my friend and longtime mentor Tim Hoctor, one of the true legends at the intersection of technology and life sciences.

Tim’s career defies categories.

He started as a professional tennis player in California. From there, he stepped into early Silicon Valley startups, then into Molecular Design Limited, the birthplace of computerized chemical registration and what many still call the “MDL Mafia.” Later, he became a senior leader at Elsevier, helping shape how scientific data, literature, and databases connect in the digital era.

This conversation is part industry history lesson, part strategic deep dive, and part personal reflection.

What we cover:

✔️ Tim’s unconventional journey from tennis pro to engineer to life sciences data executive
✔️ How MDL pioneered digital molecular representation and why it became foundational to modern pharma
✔️ Why linking structured databases to scientific literature was visionary in the 1990s and still unfinished business today
✔️ The persistent data silos in pharma and why culture, more than technology, is often the bottleneck
✔️ Why 6 billion dollar drug development costs are a systems problem, not just a science problem
✔️ The real role of regulators in AI adoption and how agencies are asking industry to help define the future
✔️ What COVID changed forever in automation, digital adoption, and supply chain resilience
✔️ How tools like ChatGPT are reshaping behavior across pharma teams
✔️ Why pharma still hasn’t had its “SpaceX moment” and what it would take to truly disrupt the model
✔️ The vision of garage biotech powered by autonomous labs, shared data, and AI driven discovery

Tim speaks with rare clarity about what holds our industry back, what gives him hope, and why better data sharing may ultimately matter more than the next algorithm.

If you care about the evolution of pharma R&D, the cultural barriers to AI adoption, and what it will take to move from incremental efficiency to true system level change, this episode is for you.

I hope you enjoy this conversation as much as I did.

🎙 Tech and Drugs – Season 02, Episode 01Inside Scientific Publishing, AI, and the Future of Medical KnowledgeIn the opening episode of Season 2, I sat down with Mitja-Alexander Linss, Head of Marketing at Karger Publishers, one of the world’s oldest and most respected medical publishers, to explore how scientific communication is evolving in the age of AI.With over 130 years of publishing history behind it, Karger sits at a fascinating crossroads: peer review, trust, and scientific rigor on one side; AI, new formats, and radically changing consumption habits on the other. This conversation dives deep into what must change, what shouldn’t, and where AI can genuinely add value without breaking the foundations of science.What we cover in this episode:✔️ How scientists and clinicians really consume content today—and why short-form, video, and audio formats are rising fast✔️ Why peer review still matters, and how new formats may (slowly) enter the “version of record”✔️ Where AI is already used in publishing: fraud detection, reviewer matching, workflow optimization✔️ Why fully automated AI peer review remains ethically off-limits—for now✔️ The data licensing debate: LLMs, copyright, fair use, and why scientific data is different✔️ How publishers can responsibly collaborate with AI companies without undermining trust✔️ Augmented and virtual reality in scientific publishing—visualizing molecules and complex data in 3D✔️ The Vesalius Innovation Award and how Karger supports startups in AI and scientific communication✔️ A forward-looking vision: from searching papers to asking questions and getting evidence-based answersIf you’re interested in AI in science, medical publishing, research integrity, or how centuries-old institutions adapt to exponential technologies, this episode offers a thoughtful, grounded perspective—without buzzwords.

🎙 Tech and Drugs – Episode #11Inside Helical: Bio Foundation Models, Virtual Cells, and the Future of AI-Native Drug DiscoveryThis week I sat down with Rick Schneider, co-founder and CEO of Helical, a Luxembourg-based startup on a mission to democratize bio foundation models.Rick and his team are building something ambitious: an open, AI-native platform that lets pharma and biotech teams actually use large DNA, RNA, and single-cell foundation models, without needing their own supercomputer or an internal army of ML researchers. From training cutting-edge mRNA models on the Luxembourg HPC MeluXina to releasing beginner-friendly open-source tools, Helical is shaping the next generation of AI for science.What we cover:✔️ Rick’s journey from “almost doctor” to engineer, AI specialist, and now biotech founder✔️ What bio foundation models really are… and why unlabeled sequencing data changes the game✔️ Why one model will never solve all of biology, and why Helical is proudly model-agnostic✔️ How the field is inching toward a “virtual cell” built by combining multimodal model embeddings✔️ Pharma’s real bottlenecks: data scarcity, batch effects, validation culture, and… organizational speed✔️ How Helical enables lab-in-the-loop workflows without operating a lab themselves✔️ The explosion of new bio models, and how Helical helps teams evaluate what’s hype vs. useful✔️ Why shifting more hypothesis testing in silico could finally compress drug discovery timelinesIf you’re curious about where bio foundation models are heading, how pharma should rethink its AI stack, or what it means to build a truly AI-native biotech platform, this conversation with Rick is packed with insights.

🎙 Tech and Drugs – Episode #10A special inside look at Ginkgo Bioworks, where automation, biology, and AI collide.I sat down with John Androsavich (GM, Ginkgo Data Points) and Jason Hocking (VP Engineering) to explore how Ginkgo generates huge multimodal datasets, automates complex biology at scale, and builds the infrastructure that today’s AI models desperately need.What we cover:✅ How Ginkgo evolved from synthetic biology to high-throughput data generation✅ Why clean, diverse datasets — not models — are the real AI bottleneck✅ A look at Ginkgo’s automation: modular “Rack” systems, NGS pipelines, and scalable cell models✅ Making data usable for both scientists and ML teams✅ The future of AI-guided labs and closed-loop experimentation✅ How shifting away from animal testing could accelerate human-relevant drug discoveryWhether you're in pharma, biotech, or AI for science, this episode is packed with insights on where R&D is headed next.

ere it is 🎙 "Tech and Drugs – Podcast" Episode #9 with 🧬💻 Douglas "Doug" Selinger from Plex ResearchThe theme for today: Applying Google's Search Recipe to Pharma R&D DataIn this episode of Tech and Drugs, I sit down with Douglas "Doug" Selinger, founder and CEO of Plex Research, whose unique approach is reshaping how scientists utilize data in drug discovery. While many companies are busy building predictive AI models, Doug and his team at Plex Research are leveraging techniques inspired by Google's internet search algorithms to make sense of vast and disparate datasets in pharma and biotech.Doug brings decades of experience, from pioneering microarray technologies in George Church's lab at Harvard to leading computational biology initiatives at Novartis. In 2017, he founded Plex Research to tackle the persistent challenges of data overload, silos, and underutilized information.We cover:✅ Doug's path from early genomics research to developing an innovative search-driven analytical platform designed specifically for scientists✅ How Plex's "focal graph" method integrates massive chemical biology and omics' datasets to reveal hidden connections in biological data✅ Overcoming Pharma’s persistent "data silo" problem by creating algorithms that adapt to the data scientists actually have✅ Real-world examples of Plex Research unlocking novel insights into disease mechanisms, biomarker identification, and precision oncology✅ Why transparency and explainability remain critical for AI adoption among scientists✅ The future potential of autonomous AI systems, combining knowledge graphs and large language models (LLMs), guided by human insightWhether you're navigating the complexities of data-rich environments or curious about pragmatic AI applications, this conversation provides actionable insights for improving decision-making in drug R&D.💬 How do you see search-inspired AI approaches changing drug discovery? Share your thoughts in the comments!