Skip to main content
Data / ML

Meet Michelangelo: Uber’s Machine Learning Platform

5 September 2017 / Global
Featured image for Meet Michelangelo: Uber’s Machine Learning Platform
Image
Figure 1: The UberEATS app hosts an estimated delivery time feature powered by machine learning models built on Michelangelo.
Image
Figure 2: Data preparation pipelines push data into the Feature Store tables and training data repositories.
Image
Figure 3: Model training jobs use Feature Store and training data repository data sets to train models and then push them to the model repository.
Image
Figure 4: Regression model reports show regression-related performance metrics.
Image
Figure 5: Binary classification performance reports show classification-related performance metrics.
Image
Figure 6: Tree models can be explored with powerful tree visualizations.
Image
Figure 7: Features, their impact on the model, and their interactions can be explored though a feature report.
Image
Figure 8: Models from the model repository are deployed to online and offline containers for serving.
Image
Figure 9: Online and offline prediction services use sets of feature vectors to generate predictions.
Image
Figure 10: Predictions are sampled and compared to observed outcomes to generate model accuracy metrics.
Jeremy Hermann

Jeremy Hermann

Jeremy Hermann is an engineering manager on Uber's Michelangelo team.

Posted by Jeremy Hermann, Mike Del Balso

Category: