Wednesday, November 17, 2021

"OPEN TALK: Using Machine Learning on Logs to Find Root Cause Faster"
Join on Hopin
Gavin Cohen
Gavin Cohen
Zebrium, VP Product

There are many ways to tell when your application breaks. But figuring out what caused it to break is slow and tedious as engineers hunt through logs and dashboards, piecing together the details of what happened.

Fortunately, unsupervised machine learning can speed-up the process. It works by automatically finding the log events and metrics that describe the root cause, and it uses GPT-3 to provide a plain language summary of the problem.