Uploaded September 2026 | Updated November 2026, 3 days ago
Heartflow's internal data platform turns data transformations into fault-tolerant, scalable pipelines running nightly on Anyscale, with no Ray code required.
At Ray Summit 2026, Deepak Sharma and Patrick Violette from Heartflow show how the platform, built on Ray Data and Ray Core, provides self-serve medallion pipelines that autoscale on spot instances, evolve schemas, and publish to Iceberg, powering everything from coronary plaque analytics to metadata extraction from millions of medical images.
You'll leave with their architecture, how they scaled it, and lessons from running production Ray across teams.
Liked this video? Check out other Ray Summit breakout session recordings: youtube.com/playlist?list=PLZNaJyYZRPdo
Subscribe to our YouTube channel to stay up-to-date on the future of AI: youtube.com/c/anyscale
đź”— Connect with us:
LinkedIn: linkedin.com/company/joinanyscale
X: https://x.com/anyscalecompute
Website: anyscale.com
Heartflow's internal data platform turns data transformations into fault-tolerant, scalable pipelines running nightly on Anyscale, with no Ray code required.
At Ray Summit 2026, Deepak Sharma and Patrick Violette from Heartflow show how the platform, built on Ray Data and Ray Core, provides self-serve medallion pipelines that autoscale on spot instances, evolve schemas, and publish to Iceberg, powering everything from coronary plaque analytics to metadata extraction from millions of medical images.
You'll leave with their architecture, how they scaled it, and lessons from running production Ray across teams.
Liked this video? Check out other Ray Summit breakout session recordings: youtube.com/playlist?list=PLZNaJyYZRPdo
Subscribe to our YouTube channel to stay up-to-date on the future of AI: youtube.com/c/anyscale
đź”— Connect with us:
LinkedIn: linkedin.com/company/joinanyscale
X: https://x.com/anyscalecompute
Website: anyscale.com










