Wednesday, October 11, 5:30pm - 5:55pm (EDT)
Your search did not return any results.
Speakers: Christopher Settles, Jean Cai, Kaavya Srinivasan
The risk knowledge platform at Uber builds and maintains the infrastructure supporting real-time machine learning use cases across fraud, identity, safety, and compliance for every vertical Uber operates. This wide adoption is made possible via the many systems they maintain, such as their graph database, streaming feature computation platform, and tools they build to enable point in time correct joining for features for machine learning training data construction.
In this talk, they dive into the scale & architecture of this infrastructure, and shed light on the specific use cases for why their custom solutions are required for tackling the challenging problems they face across the risk domain.