Really happy to have been a part of the team behind the Test at Scale platform!

LambdaTest Test at Scale is an intelligent test execution platform that cleverly selects the specific tests to run based on the changes made during a specific commit! This significantly speeds up your software testing pipeline and enables software teams to achieve a high velocity!

Currently TAS supports JavaScript testing frameworks like Jasmine, Jest and Mocha. It also supports Java, with more languages and frameworks coming soon :)

Check out their repository at https://github.com/lambdatest/test-at-scale

looks interesting. Do they support flaky test detection?

Instead of flaky test detection, TAS supports proactive flaky test management where in for every github PR, TAS will flag the flaky tests so that appropriate decision can be taken

I don't understand why the instructions ask to create an account. Does the repo in GitHub has the main logic of the optimization, or it's just a thin wrapper that calls the API in your server?

Disclaimer: I belong to the development team behind the TAS. The github repo is a runner which uses APIs to send test metadata to server. Account creation is required to associate and store meta data to provide test intelligence data.

Does TAS work on On-Prem cloud? Also, what magic TAS weaves behind the scenes so that we get the most of the CI pipeline?