At the 2025 @Deloitte AI Forum, Pete Shimer and Brian M. Bot chatted about critical role of data federation for unlocking progress in cancer research. In our latest blog post, we offer 4 takeaways from Pete and Brian’s conversation: https://t.co/SMQ2PeRjgt

CAIA places AI at the intersection of medicine, science, and data, enabling cancer centers to focus on generating new insights and advancing cancer discovery.

CAIA's standardization process includes: ➡️ Addressing the challenge of differing data structures and formats. ➡️ Adopting the OMOP Common Data Model (CDM) to map local codes into a standardized system. ➡️ Ensuring all researchers only access de-identified clinical data.

Building powerful, multi-institutional AI models for cancer research requires a common data language. In its first year, the CAIA team set out to solve this data standardization challenge. Read more:

.@CAIAorg launched a federated learning platform and will pilot Asta DataVoyager, an @allen_ai data analysis AI agent, to ask questions of datasets in plain language. https://t.co/2dVfO8tgPa via @hitanalytics

Many thanks to our foundational cancer centers including @DanaFarber, @MSKCancerCenter, @fredhutch, @HopkinsMedicine and @HopkinsEngineer

This progress includes: ➡️ A new legal framework and successfully navigating regulatory pathways across all 4 participating centers ➡️ A Gen 1 dataset comprising structured clinical data from over 1 million patients ➡️ 8 pilot use cases to prove the platform's immediate impact

Earlier this month, we announced our collaborative AI platform at @MadronaVentures 's IA summit. This launch is a reflection of the truly collaborative ethos upon which CAIA was built. Read our blog post on our progress in just one year: https://t.co/Ng4bh2CN0d

This framework allows CAIA to scale cancer research by combining collective insights from diverse datasets, leading to more accurate and equitable AI models, all while strictly maintaining data privacy.

2. It acts as an aggregator, receiving summarized, de-identified model updates from each center, which are then combined to create an updated global model. 3. Crucially, the orchestrator never accesses sensitive patient data, which is secured behind institutional firewalls.

A quick summary: 1. The orchestration layer leverages @nvidia FLARE to coordinate the federated learning network, securely distributing model code and instructions to local centers.

📣 New blog post alert! In our latest post, we take a closer look at how CAIA's orchestration layer securely enables AI insights in a federated learning framework. https://t.co/UHfWQVigcF

Can #AI really cure cancer? Fred Hutch Vice President & Chief Data Officer Jeff Leek explains how @CAIAorg is breaking down research silos with federated learning, accelerating the pace of research breakthroughs. Listen to more from @KUOW: https://t.co/5chVyxcJvD

+1 poised to share our scientific discovery efforts @allen_ai with Cancer AI Alliance (@CAIAorg) to move the needle Blog: https://t.co/cUk0hstpzC AI + cancer research + open-source More coming soon

Can #AI really cure cancer? Fred Hutch Vice President & Chief Data Officer Jeff Leek explains how @CAIAorg is breaking down research silos with federated learning, accelerating the pace of research breakthroughs. Listen to more from @KUOW: https://t.co/5chVyxcJvD

This platform allows researchers to leverage collective insights rather than working in isolation, significantly accelerating our goal to improve clinical outcomes for cancer patients. Read more about our approach to collaborative action in Newsweek:

Last week, CAIA announced the first collaborative, scalable platform using federated learning for cancer research. With federated learning, researchers can accelerate discovery while maintaining the highest standards of data security and privacy

Cancer AI Alliance @CAIAorg debuts #cancer #research platform - enables participating cancer centers to train #AI models on millions of clinical data points without moving individual data outside institutional firewalls

CAIA is going places this week — specifically to the @NYSE trading floor to celebrate the launch of our collaborative AI platform for cancer research! We are grateful for this recognition and support for our mission to accelerate cancer research and discovery.

“As part of this alliance, we are moving beyond generic applications of AI and building world-class, state of the art AI systems that could actually comprehend the profound complexities inherent within the cancer domain” - Ai2 CEO Ali Farhadi on @CAIAorg