Uber Engineering(@UberEng):How do you keep ML features consistent at 8M QPS? We built a feature logging framework that selectively captures the features used at inference for training, reducing training-serving skew without logging trillions of unnecessary rows. See how it works. ⬇️

Uber Engineering

@UberEng

Engineering @Uber, from practice to people. ✨

Joined June 2014

5.4K Following 71.1K Followers

Uber Engineering@UberEng

2026.09.22 17:00

How do you keep ML features consistent at 8M QPS? We built a feature logging framework that selectively captures the features used at inference for training, reducing training-serving skew without logging trillions of unnecessary rows. See how it works. ⬇️