Why AI-Native Cell Manufacturing Is the Future of Drug Discovery (Part 1)

*Full transcript available below

Ahead of the 2026 J.P. Morgan Healthcare Conference, occurring Jan. 12–15 in San Francisco, Daniel Delubac, PhD, CEO and co-founder of US biotech firm iOrgan Bio, sat with BioPharm International® to discuss the impact of digitalization on cell engineering. Dr. Delubac is betting the future of drug development on AI-designed human-derived models.

He argues that today’s system is fundamentally broken, pointing to the roughly 95% of drug discovery campaigns that fail, largely because they rely on animal models—most often mice—that do not truly represent human biology. While the ethical case for phasing out animal testing is clear, Dr. Delubac emphasizes the utilitarian case, asserting that replacing animal models with precise, human-derived systems is the most important lever to increase success rates in drug development.

How can AI-driven digitization replace unreliable animal models in drug development?

Dr. Delubac’s vision centers on digitization and AI to define, design, and manufacture human-derived cell models. First, data-driven methods create a “CAD [computer-aided design]-like” digital specification of the ideal cell state—strict requirements and features that serve as a standard reference, he explains. Then, AI-driven manufacturing dynamically monitors and adjusts live cell processes to hit the digital target, despite the inherent variability of living systems and unique human biology, he adds.

“We need digitization and AI to successfully be able to create consistent products at the end of the manufacturing process,” Dr. Delubac states.

The aim is to achieve nothing less than consistent, high-fidelity human models that can reliably predict clinical outcomes and finally decouple drug testing from animals.

Dr. Delubac will be giving a talk at the conference at 11:00 AM ET on Tuesday, Jan. 13.

Click here for more conference coverage.

About the speaker

Daniel Delubac, PhD, CEO, iOrganBio

Dr. Delubac is a technical leader with a strong track record in building transformational laboratory products and techbio platforms across AI, software, robotics, and process engineering. He has built some of the most impactful healthcare products in non-invasive prenatal testing at Counsyl, liquid biopsy at Guardant Health, early cancer detection at Freenome, patient-derived models of cancer at Xilis, and the first AI-agentic synthetic chemistry platform at Chemify. Dr. Delubac trained as an applied physicist, then earned a PhD in Biomedical Engineering from Carnegie Mellon University. At iOrganBio, he is rethinking cell-based products from an engineering and data-driven perspective and building the next generation, AI-enabled infrastructure designed to manufacture the exact same cells we have in our body—that will model biology in vitro accurately, restore degraded bodily function, cure disease, and power population-scale, affordable, holistic, regenerative medicine.

Transcript

Editor's note: This transcript is a direct, unedited rendering of the original audio/video content. It may contain errors, informal language, or omissions as spoken in the original recording.

Speaker 1

I am Daniel Duluback. I am currently serving as the CEO of iOrganBio. I hold a PhD in Biomedical Engineering from Carnegie Mellon University. I've spent 10 years in Silicon Valley building three Bleeding Edge diagnostic companies that would be Council, garden, health and freedom.

Then I've built Xilis, a platform company focused on the creation and screening of patient derived model of cancer. Most recently, I was in the UK building Chemify, a deep technology company digitizing and automating synthetic chemistry.

And I've started working at I organ bio as its cofounder and CEO at the beginning of 2025 with professors Shubin Shen and Shilling Shen.

So first, I'd like to say that we're thrilled to be part of this long overdue movement to phase out animal testing for ethical reason, but also from a utilitarian perspective, we expect the human derived model will be the single most important factor in increasing the rate of success in drug development. Currently, 95% of drug discovery campaigns fail because we select drugs using animal testing.

Hence, select drugs that are safe and effective for animals, most often mice, and they end up not working in humans. Because, believe it or not, we are not large mice. The process of developing more accurate animal models, more precise, more representative models of human biology, in 20 years since the development of the reprogramming factors by Professor Yamanaka have not yet led to models that are perfectly representative of human biology without AI and without data driven approaches, we will not be able to reach this expected state of being able to recreate accurately and precisely human biology In the laboratory, we think of it as a critical prerequisite to the achievements of this vision to successfully replace animal models with human derived one manufactured in the laboratory.