In a recent episode of the Startup and Scaleup People podcast, our Founding COO, Mathilda Strom, sits down with host Mario Rivas to discuss how we are leveraging AI to fundamentally change drug discovery and address the pharmaceutical industry's 90% clinical trial failure rate.

Currently, drug development is a costly 7-to-12-year process where the majority of funds are spent on clinical trials that frequently fail due to poor patient selection. While many AI applications focus strictly on early-stage molecular design, Mathilda explains that our mission at Bioptimus is to build foundation models that understand patient biology holistically. By predicting how a patient will be affected by a drug in their entirety, we can help pharmaceutical companies weed out likely failures and drastically improve clinical success rates.

Drawing on her previous experience scaling health and insurance access to over 100 million users in emerging markets with Beimma, Mathilda also details the core business strategies driving our approach. She emphasizes that even the most complex AI models must prioritize simplicity, operational attention to detail, and "last-mile adoption" to ensure that the healthcare industry fully understands and utilizes the tools we are building.

In this comprehensive interview, Mathilda explores the core challenges and the exciting future of our field, highlighting:

• Fixing Clinical Trials: The pharmaceutical industry's biggest opportunity for AI is addressing the 90% clinical trial failure rate. Two-thirds of pharma R&D spend goes into these trials, which often fail due to poor patient selection. We are building models that understand patient biology holistically to help solve this.
• Scaling and Simplicity: Drawing on her experience scaling healthcare to over 100 million users in emerging markets with Bima, Mathilda highlights that a founder mindset focused on simplicity is what truly drives scale.
• Operational Insights: Scaling requires intense attention to detail. Mathilda explains why tracking every unit cost is necessary to win at scale, and why "last-mile adoption" is where true growth happens.
• The Future of Scientific Work: AI will not replace scientists; it will elevate them. By automating mundane administrative tasks, AI will allow scientists to transition into "editors" who focus on high-impact, meaningful questions.
• AI Governance: As we look toward the future, Mathilda shares her governance insight that building global AI infrastructure requires robust private innovation combined with appropriate government regulation.

‍