Skip to main content

The Uber Engineering Tech Stack, Part I: The Foundation

19 July 2016 / Global
Featured image for The Uber Engineering Tech Stack, Part I: The Foundation
Screenshots show Uber’s rider app in New York, China, and India as of spring 2016.
In April 2015, Uber’s 300 operational cities were scattered across the map.
At the end of summer 2014, refactored the system to match this high-level architecture.
We compare dynamic thresholds (red, yellow) generated hourly from our outlier detection algorithm to the in-streaming data. We’ve added the actual minute-by-minute data for reference (blue), which is of course unknown ahead of time. The thresholds closely follow the pattern of the actual metric.
Table and confidence interval visualizations add functionality to our A/B testing platform, Morpheus.
Visualizations demonstrate the capability of , a wrapper for developed by the Uber Data visualization team.
An Uber heatmap shows the density of vehicles driving passengers. We then remove the top percentiles from the distribution to see the underlying one.

Posted by Lucie Lozinski