AI and Statistics Start-ups: Opportunities and Challenges

In this thought-provoking keynote recording, Manjari Narayan takes us on a journey through one of the most pressing and promising intersections in modern science: the convergence of artificial intelligence, statistics, and biotechnology. Drawing on her extensive experience in both academia and biotech startups, Manjari explores the critical role statisticians can play in AI-driven drug discovery, biomarker validation, and experimental design. We are living in a "Cambrian explosion" of biotechnology, where high-throughput experiments, protein engineering, humanized models, and AI-powered screening open massive opportunities—but also introduce challenges in scientific validity, reproducibility, and decision-making. Through personal vignettes and cutting-edge examples, Manjari lays out how statistical thinking can (and should) drive better outcomes in early-stage drug development, biomarker discovery, and translational model evaluation. This episode is a must-listen for statisticians, data scientists, and healthcare innovators navigating the rapidly evolving biotech and AI startup landscape.

In this special keynote episode, I’m excited to share the recording of Professor Tim Friede’s thought-provoking presentation from The Effective Statistician Conference 2024. Tim, a leading expert in biostatistics and clinical trial design, dives deep into the combination of randomized controlled trials (RCTs) and real-world data (RWD)—especially in the context of rare diseases. Drawing from his work at the University Medical Center Göttingen and numerous European research initiatives, Tim presents a compelling case for integrating RWD to support small or underpowered RCTs using advanced statistical models. He shares real-world examples (including CJD and Alport syndrome), simulation insights, and practical recommendations that can change how we approach evidence generation in low-prevalence populations. What You’ll Learn:

In this keynote episode, Professor Sebastian Schneeweiss from Harvard Medical School shares groundbreaking insights from his extensive research into emulating randomized controlled trials (RCTs) using real-world data (RWD). Recorded live at The Effective Statistician Conference 2024, this talk explores whether non-randomized studies based on electronic health records and claims data can reach conclusions as reliable as those from traditional RCTs. Prof. Schneeweiss, also Chief of the Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women’s Hospital, walks us through the RCT DUPLICATE project, a major FDA-funded initiative that evaluated whether regulatory decisions could be replicated through high-quality real-world evidence (RWE). From the successes to the limitations—and everything in between—this episode is packed with lessons for statisticians, regulators, and pharmaceutical leaders interested in the future of data-driven healthcare decisions.

In this episode of The Effective Statistician, Alun Bedding takes the mic to explore one of the most underrated yet transformative tools in any professional's career—networking. With the PSI Conference happening, this episode offers timely insights into how statisticians—introverts and extroverts alike—can use networking to build meaningful professional relationships, accelerate their careers, and find support within the broader community. Alun shares personal anecdotes, including how a casual conversation with Alexander at a past PSI event led to his current guest host role, and outlines a step-by-step mindset and strategy for networking with authenticity and purpose.

In this episode, Alun Bedding welcomes Emma Crawford for a powerful and personal conversation about building inclusive workplace cultures—ones where every individual can thrive. Emma shares her late diagnosis of autism and how it reshaped her experience and expectations of the workplace. Together, they explore the limitations of the traditional “reasonable adjustments” framework and introduce the concept of success enablers—workplace strategies and tools that benefit not just neurodivergent individuals but everyone on the team. The conversation dives deep into the role of leadership, AI tools, flexible work arrangements, and cultural shifts that prioritize accessibility and well-being for all employees. Emma also previews her interactive workshop at the upcoming PSI conference and gives insight into how the session will encourage open discussion, hands-on activities, and actionable takeaways to help leaders and teams foster a more inclusive environment. Whether you’re a people leader, statistician, or advocate for change—this episode will challenge your assumptions and inspire more inclusive practices.

How can oncologists and healthcare professionals keep up with the ever-growing body of research to make the best decisions for patients? In this episode, I speak with Anna Forsythe, a pharmacologist, health economist, and founder of OncoScope, a groundbreaking platform delivering daily updated systematic literature reviews (SLRs) in oncology. Drawing on decades of experience in pharma and health economics, Anna shares how automation and AI are transforming the traditionally tedious SLR process—making up-to-date evidence accessible to clinicians in just a few clicks. Anna’s vision is clear: democratize access to high-quality, current evidence for clinicians—and ultimately improve patient care.

In this episode, I’m joined by Julia Geronimi from Servier and Dr. Pavel Mozgunov from the University of Cambridge to explore a topic that’s absolutely central to advancing precision medicine—predictive biomarkers. We dive into the challenges of identifying predictive vs. prognostic biomarkers, especially in early-phase clinical trials with limited sample sizes. What makes their approach so exciting is that it offers a model-flexible, visually intuitive way to detect predictiveness—even before we talk about dichotomizing biomarkers or setting cutoffs. If you work on clinical trial design, translational science, or biomarker development, this conversation will give you fresh tools—and a lot to think about.

In this special episode, I’m sharing the recording of a webinar I co-hosted with Cytel on March 20, 2025. I was joined by an expert panel of leaders in statistics and clinical development: Yannis Jemiai, Flaminia Chiesa, and Benjamin Piske. Together, we explored how the role of statisticians is rapidly evolving in response to industry changes, data innovations, and AI-driven transformation. This rich discussion dives into what it means to lead as a Clinical Data Scientist today—and why statisticians are uniquely positioned to influence strategy, innovation, and decision-making across the healthcare and pharmaceutical sectors.

In this episode, I’m excited to welcome back Katrin Kupas, a statistician with deep expertise in Health Technology Assessment (HTA) and real world evidence (RWE). We dive into how RWE and the new Joint Clinical Assessment (JCA) process in Europe can work together—and where the challenges lie. As the JCA becomes more central in EU regulatory and reimbursement discussions, knowing how and when to use real world data is critical. Katrin shares practical use cases, methodological guidance, and strategic insights for integrating RWE into early planning.

As someone who found my way into statistics late—only halfway through university—I often wonder: what if more young people knew about this path earlier? In this episode, I sit down with two wonderful guests and PSI volunteers, Emma Crawford and Alex Spiers, to explore exactly that: how we, as statisticians and scientists, can inspire the next generation. We talk about the why behind investing in STEM outreach, share personal stories, and get into the practical steps you can take—whether you want to volunteer at a school, present virtually, or simply start a conversation with a student.