Unleashing Michelangelo’s power for Uber’s AI transformation.


TLDR:

  • Michelangelo plays a key role in Uber’s ML evolution since 2016
  • Approximately 400 active ML projects are managed on Michelangelo, with over 20K model training jobs monthly

Article Summary:

The article highlights Uber’s journey in AI and ML, emphasizing the role of Michelangelo in accelerating Uber’s AI evolution since 2016. It covers the evolution of AI/ML at Uber, the three phases from 2016 to now, and the introduction of Generative AI. The article discusses the importance of the ML Developer Experience, the journey of Michelangelo 1.0 to Michelangelo 2.0, and the introduction of Generative AI in 2023. It also delves into the architectural overview, model quality, project tiering, model iterations as code, deep learning integration, and MA Studio for ML development. The article concludes by discussing Uber’s investment in Generative AI and the enhancements made to Michelangelo to support LLMOps and GenAI applications.