Protege

The availability of training data is the biggest unmet need in AI. After hundreds of conversations around training data, I've put together the patterns that are most commonly holding back data deals today https://lnkd.in/eEg8KWUw

Last week, our CEO Bobby Samuels?sat?down with Sarah M. London, CEO of Centene Corporation, to moderate the keynote discussion at our AI in Healthcare Summit. Check out a summary of that discussion here. Big thanks to our co-hosts Shaper Capital and Fractional AI for making the Summit such a success! https://lnkd.in/e5Q4w7u3

Announcing our partnership with HistoWiz, which will bring their extensive digitized pathology datasets to AI builders in the Protege network! Looking forward to working together to make this valuable data available to AI innovators. Learn more here: https://lnkd.in/eTJTHjNN

A Rubric for Evaluating Healthcare AI Training Data In the fast-evolving world of AI, bigger data isn’t always better. Simply scaling the volume of training data won’t be enough to drive the next generation of models. Instead, AI is shifting toward more specialized models, where the quality and diversity of data play a more critical role than ever before. ?? So, how do we evaluate if a dataset is AI-ready? In our latest post, we outline 5 key dimensions to assess training data quality: ? Connectedness of patient journeys – Does the data reflect real-world care pathways? ? Representation of diverse subpopulations – Are all key demographics included? ? Richness of positive cases – Does the dataset contain sufficient variation? ? Quantification of dissimilarity – How do we measure dataset diversity? ? Multimodal capture – Does the data reflect all sources used in clinical decisions? As AI models become more specialized, data curation becomes a multi-constraint optimization problem—balancing tradeoffs between completeness, diversity, and real-world applicability. ?? Read the full post for a structured framework on evaluating healthcare AI datasets: [https://lnkd.in/ejB82F3m] What challenges have you faced in curating high-quality training data for AI? Let’s discuss in the comments! ?? #HealthcareAI #DataCuration #AITraining

The National Academy of Television Arts & Sciences hosted an AI symposium this weekend featuring Protege's GM of Media, Dave Davis, on a panel discussing the practical considerations and implications of licensing content to AI companies for model training. Variety’s Carolyn Giardina covered the highlights here: https://lnkd.in/gvz_22aS Eric W. Shamlin Seth Hallen Ed Ulbrich Barbara Ford Grant Holly Leff-Pressman Erik Weaver Jason Zada Stephen Fefferman Carlos Sanchez Xenia Shevnina Ali Hekmatpour James Golden Jim Tosney Han Seng Lim Pablo Iacoviello Mirjam Laux Wendy Bernfeld Max Einhorn Brendan Gallagher Robyn Polashuk Andrew Folks Julie Shapiro

???Mapping the Future of Healthcare AI??? AI is transforming healthcare—enhancing diagnostics, streamlining operations, and improving patient outcomes. But with so much innovation happening, understanding the landscape can be overwhelming. That’s why?Protege?created the?Healthcare AI Market Map—a comprehensive breakdown of the AI-driven solutions shaping the industry. We've categorized the key players across three major areas: ???Providers & Patients?– AI-powered diagnostics, patient engagement, and clinical decision support. ???Life Sciences?– AI in drug discovery, clinical trials, and protocol optimization. ???AI Infrastructure?– The backbone powering AI-driven healthcare solutions. This is your go-to guide for understanding the evolving AI ecosystem in healthcare. Check it out now! https://lnkd.in/erEgX99D

Not every day you have an industry titan analyze your business... many thanks to?Nikhil Krishnan?who laid out our key collaborative dynamics with AI builders and data holders in healthcare with his iconic insight and humor. Check it out: https://lnkd.in/gXHT65NN

Announcing our partnership with Shaip! Together, we're?bringing a diverse set of healthcare and medical data to responsible innovators in the AI market via Protege's platform. Read more here:?https://lnkd.in/eA4KTfEk

Our healthcare vertical continues to grow, and we're hiring in our data science area! As an Applied Data Scientist: ???? work at a fast pace; ?? with people who are super sharp *and* super kind; ?? helping AI builders craft the right dataset from the largest training data repository available in healthcare today. Application link here: https://lnkd.in/ekWY6Wcc

Announcing our partnership with iCliniq, which will bring their extensive physician-audited real-world clinical datasets to AI builders in the Protege network! Looking forward to working together to make this valuable data available to AI innovators. Learn more at the link here: https://lnkd.in/ePucdDQ7