Talk: Smart scrubbing of test results

We use ML clustering to identify similar logs to detect same type of failures with the goal of scrubbing test results a lot faster.

By extracting features from unstructured log files we can use clustering techniques to put results in buckets we can scrub together.