SYSTEM Cited by 1 source

Michelangelo (Uber ML platform)

Michelangelo is Uber's end-to-end internal machine learning platform — feature engineering, training, model management, and serving for all of Uber's ML-driven products (dispatch matching, pricing, fraud, Eats recommendations, ETA prediction, driver incentives, etc.). (Source: sources/2024-03-14-highscalability-brief-history-of-scaling-uber)

Primary-source reference: Scaling Michelangelo.

Stub page — expand on future Michelangelo-internals ingests. Michelangelo is the ML infrastructure counterpart to Uber's data-infrastructure arc (Hudi + Presto on Kafka + Spark), completing the end-to-end data-→-model pipeline.

Seen in

• sources/2024-03-14-highscalability-brief-history-of-scaling-uber — cited as the ML platform built at Uber scale, alongside the data-infrastructure systems (Hudi, Presto, Spark) in the "other large initiatives" section.

Related

• companies/uber — origin org.