The Bits in Bio Podcast

Akhila Kosaraju, MD, is the co-founder and CEO of Phare Bio, a nonprofit biotech using generative AI to discover entirely new classes of antibiotics, built in partnership with Jim Collins’ lab at MIT, the team behind the landmark AI-discovered antibiotic Halicin.We discuss the leap from predictive to generative AI in antibiotic discovery, why synthesizability remains the biggest unsolved bottleneck, and how Phare’s open-science model is sharing data to lift the entire field at a moment when only 3,000 researchers worldwide are working on antibiotics.From the Pentagon to raising from TED’s Audacious Project, ARPA-H, and Google, Akhila shares why the antibiotic market is broken, what a subscription-based pull mechanism could fix, and how she leads a flat, multilingual team at the intersection of AI and infectious disease.

Sid Sijbrandij, co-founder and Executive Chair of GitLab, and Jacob Stern, formerly of 10x Genomics, join us for our first-ever live episode to share how Sid went “founder mode” on his aggressive bone cancer after running out of standard treatment options.We discuss the four principles behind their approach: maximal diagnostics, creating treatments, running treatments in parallel, and scaling for others. Jacob explains how single-cell sequencing guided an experimental radioligand therapy, and how a scan in Beijing reshaped a personalized cell therapy.From publishing 25TB of Sid’s tumor data openly to launching twelve companies through Even One Ventures, Sid and Jacob share lessons on rescuing promising drugs, faster regulatory pathways, and making this approach accessible to everyone.

Dov Gertz, co-founder and CEO of Converge Bio, is building a generative AI lab for the life sciences, with AI systems that plug directly into drug development workflows, from target discovery to antibody design to manufacturing.We discuss why generative AI in biology needs systems rather than single models, where de novo design is overhyped, and how Converge's antibody system layers physics-based docking on top of generative models. Dov also explains how on-demand biobetters could reshape an industry built on patent expiration.From discovering novel CRISPR systems with machine learning to building Converge across Tel Aviv and Boston, Dov shares lessons on following your passion, building talent density, and why engineers entering bio need more biology than they think.

Michelle Lee, PhD, is the founder and CEO of Medra, building a Physical AI Scientist platform where robotic agents and reasoning models run, optimize, and learn from wet lab experiments together.We discuss Medra Lab 001, the largest autonomous life sciences lab in the US, built in just 90 days, and the Vision-Language-Lab-Action model that powers it. Michelle breaks down Medra's "data foundry" thesis for foundation model teams, and why she made the deliberate bet to build in America after decades of automation work moved offshore.From a robotics PhD at Stanford to running production science for Genentech, Michelle shares why she believes AI won't transform drug discovery until it leaves the screen, and what it actually takes to build and fund a deep-tech, full-stack biotech from scratch.

Pauric Bannigan, co-founder and CSO of Intrepid Labs, is building one of the first AI-native platforms for drug formulation, combining machine learning, robotics, and experimental science to tackle one of pharma’s most complex bottlenecks.We discuss why drug formulation remains a critical but underexplored step in drug development, and how Intrepid’s “self-driving lab” integrates AI-driven decision-making with automated experimentation to accelerate formulation design.From academic research in pharmaceutical chemistry to founding an AI-first biotech company, Pauric shares lessons on building at the intersection of machine learning, robotics, and drug development, scaling interdisciplinary teams across computation and experimental science, and transforming formulation from a manual bottleneck into a data-driven system.

Molly Gibson, Origination Partner at Flagship Pioneering, has co-founded some of the most ambitious AI-native life sciences companies of the last decade, including Generate Biomedicines, Lila Sciences, and most recently Expedition Medicines, where she serves as CEO.We discuss the Flagship model of company origination, how Lila Sciences is building AI Science Factories where AI runs the scientific method autonomously, and how Expedition Medicines is using AI and quantum chemistry to crack the undruggable proteome.From her PhD studying antibiotic resistance in preterm infant microbiomes to founding multiple AI-bio companies, Molly shares hard-won lessons on separating AI hype from real platform value, building cross-disciplinary teams, and what computer scientists need to understand before becoming biotech founders.

Chris Petersen, co-founder and CTO of Scientist.com, joins us to share lessons from almost two decades of building the world's leading research services marketplace for pharmaceutical R&D.We discuss pivoting from building a CRO lab to creating a B2E marketplace platform facilitating $2.5B in research procurement, the critical architectural decisions along the way, and deploying AI to streamline workflows for nearly all top 30 pharma companies.From a legendary Paul Blart Mall Cop rescue in the early days to recent acquisition by GHO Capital Partners, Chris shares hard-won insights on building for enterprise, hiring through community, and finding opportunities hidden in Excel spreadsheets.

Frank Li, founder and CEO of Stately Bio, joins us to unpack how AI-driven, label-free live-cell imaging is unlocking real-time insight into cell state and redefining what’s possible in regenerative medicine.We discuss his journey from co-founding a mobile payments startup through roles at Palantir, Google Brain, and Calico, to building Stately Bio’s ML imaging platform, and how surprising predictive power emerges when you stream and analyze living cells at scale.From three years in stealth to a $12M seed round, Frank shares lessons on bridging tech and biotech, building a world-class technical team, and applying AI to one of biology’s most complex frontiers.

Armand Cognetta, founder and CEO of General Proximity, joins us to explore how proximity medicine is revolutionizing drug discovery by targeting proteins once considered impossible to drug. We discuss the evolution beyond PROTACs, being a solo bio-founder in Y Combinator, and Armand's journey from imposter syndrome to raising $16M and winning a $3M ARPA-H grant.From nearly running out of money to building breakthrough platform technology, Armand shares hard-won lessons on fundraising, team building, and maintaining startup speed in an industry known for decade-long timelines.

Robin Roehm, CEO and co-founder of Apheris, joins us to explore how federated learning is transforming data collaboration in life sciences. We discuss its impact on drug discovery and clinical trials, the balance between IP protection and AI innovation, and Robin’s journey at the crossroads of tech and healthcare.