Kevin G. Libuit

Science. Technology. And Business. That’s the banner I run under with Down to a Science. And @Niamh_Oh and diagnostics. The technology to better characterize pathogens and serve patients. And the business of navigating insurance, regulations, and clinical adoption. This one was a blast. We covered Niamh's journey from PhD to CEO and had a chance to get surprisingly deep into the technical details of microbial characterization with AI (not the scary kind!) and metagenomics. My personal takeaways: - PhD Candidates and Start Ups Founders have a shared archetype: Both self-driven actors who will bring ideas to fruition on shoe string budgets and little guidance. Also, PhD candidates have an advantaged perspective to spot technical arbitrages that a start-up can take advantage of - Getting sequencing into hospitals is a long game: The technology is advancing rapidly, but clinical validation, regulation, and reimbursement remain major hurdles. Who actually pays for these tests is still a difficult question. And I really appreciate @BiotiaInc's approach here: sequencing for diagnostics feeds into a larger system of biosurveillance. Two birds fed with the same scone. - More data doesn't automatically mean better care: Metagenomic sequencing can identify pathogens conventional methods might miss. But translating those findings into actionable clinical information is the hard part (and also validating it and getting it cost-optimized). A fantastic conversation about what it takes to move science from the lab into clinical practice. Thanks to Niamh for joining me! youtube.com/watch?v=oWOeYgaY…

At what point is someone technical?

If you're in Bioinformatics, @tangming2005 is a voice worth following. He's a titan in the field who's helped thousands (myself included) learn the technical foundations of our craft. And in this episode, we talk all things AI and its impact on bioinformatics: What it can do, what it can't do, and what responsible implementation looks like. Unsurprisingly, he's come to a similar conclusion as specialist across other disciplines: The barriers to technical execution have lowered dramatically, meaning that the real value has shifted toward knowing what problems to solve, how to frame them, and defining metrics to judge results. Full conversation here: youtube.com/watch?v=zm1wIXKL…

An industrial engineer, product manager, and applied AI (SI?) specialist walk into a bar... Turns out its all @brandon_galang. A guy who's held many hats in his career that have all helped him build a powerful perspective on how to effectively identify and solve problems. He calls it an engineering approach. I call the scientific method: Assess an environment, consider how variables relate, test it. Repeat. From writing content for his Substack, building AI solutions, and even job hunting, Brandon takes this same approach to break down complex challenges, test assumptions, and turn ideas into actionable solutions. Such a great convo. youtube.com/watch?v=VOC3PHUh…

Sandeep Pawar is a self-proclaimed data nerd. So he was in good company on the DTAScience Podcast 🤓 My main takeaways from our conversation: 𝗦𝗼𝗽𝗵𝗶𝘀𝘁𝗶𝗰𝗮𝘁𝗲𝗱 𝗱𝗮𝘁𝗮 𝗿𝗲𝗾𝘂𝗶𝗿𝗲𝘀 𝗮 𝘀𝗼𝗽𝗵𝗶𝘀𝘁𝗶𝗰𝗮𝘁𝗲𝗱 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 - Your data is your gold and trusted research environments are built to protect it well while enabling the appropriate analysis by the appropriate personnel. 𝗘𝘃𝗲𝗻 𝗶𝗻𝘁𝘂𝗶𝘁𝗶𝘃𝗲 𝗔𝗜 𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻𝘀 𝗯𝗲𝗻𝗲𝗳𝗶𝘁 𝗳𝗿𝗼𝗺 𝗮𝗻 𝗼𝗻-𝗿𝗮𝗺𝗽 - There's are reason that FDEs are the fastest growing job in AI. This technology can accelerate anything; translating that potential into worth-while outcomes takes an engineering mind that knows how to piece the puzzle effectively. 𝗠𝗲𝘀𝘀𝘆 𝗱𝗮𝘁𝗮 𝗶𝘀 𝗮 𝘂𝗻𝗶𝘃𝗲𝗿𝘀𝗮𝗹 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 - Biopharma's current challenge with RWD is another iteration of a familiar story. AI doesn't remove the responsibilities of expert design, standardization, and secure integration. youtube.com/watch?v=KJeHyHac…

This one is about Service, Science, and Art. Dr. Tshaka Cunningham is a molecular biologist and biotech entrepreneur who has spent decades investing into his community. In this conversation, Dr. C shares how the three most influential women in his life (his mother, grandmother, and great grandmother) helped form the values and curiosity that still guide his work today, including his commitment to fostering opportunities for the next generation. A very special conversation that I’m excited to share: youtube.com/watch?v=3J96ZHyc… ----- Dr C. and I got connected through his work at Athari BioSciences and their LIFT-VA program. We're working alongside the Athari team to help grow the Virginia life science workforce by giving students and emerging professionals access to the kinds of opportunities that often feel out of reach. If you're at all interested in this type of work, please reach out.

AI adoption is a big deal in any industry. Public health more so than most. The food we eat, water we drink, and the air that we breathe all depend on the scientists, systems, and infrastructure aimed at promoting and protecting the public health of our communities. Hard to find higher stakes than this. So when a transformative technology like AI is brought to the table, public health leaders face a difficult balancing act: Move too quickly and you risk undermining the quality, trust, and accountability these systems require. Move too slowly, and you risk leaving critical gaps unaddressed when the tools and capabilities to save lives and prevent disease are available to your communities. In this episode, Luke Short, Director of the Dallas County Public Health Laboratory, talks about leaning into that tension. He talks about why public health needs to embrace AI, how it can expand the capacity of teams operating with fewer resources, where humans must remain in the loop, and how AI can be used responsibly even inside highly regulated laboratory environments. I’ve spent the past decade working toward bioinformatics capacity development across public health. Bioinformatics was a transformational technology that fundamentally changed how laboratories operate and how we conduct pathogen surveillance. And there are important lessons in that experience for the current challenges around AI We know that adopting new technology is more than jsut making it available. Responsible adoption happens when it can be integrated into existing systems, trusted by the people using it, validated for the work it supports, and backed by a workforce prepared to use it well. So with AI, we have an opportunity to apply those same lessons. To move with enough urgency to capture what this technology can offer, while building the systems, safeguards, and workforce needed to use it well. Check out our convo here: youtube.com/watch?v=j0jBuQr6…

Homework has always been a poor proxy for comprehension. AI made it worse.

This is a sound and practical push back on a lot of the promise of AI., especially as it pertains to speed in regulated fields. The asesment of impact can't be done in silico.

Alex Sundermann is a scientist I’ve enjoyed learning from. In particular, I’m fascinated by his perspective as a boots-on-the-ground infection preventionist who now works to integrate NGS and bioinformatics in healthcare. His work highlights the undeniable impacts of these technologies. Both clinical and economic. But pathogen genomics in healthcare remains the exception, not the norm. We spent an hour exploring the challenges holding the field back and–more importantly–proposing solutions that could help to push things forward. Full episode is live. Let us know your thoughts! youtube.com/watch?v=XDLokj8D…

Great opportunity to join a leading laboratory in our field