Solution

Send developer productivity to the stratosphere

Launchable’s core technology is a machine learning engine that predicts the likelihood of a failure for each test case given a change in the source code. This allows you to run only the meaningful subset of tests, in the order that minimizes the feedback delay.

Today, most software projects simply run all the tests all the time, in no particular order. When you are working on a small change in a large project, this is incredibly wasteful. Intuitively, you know that only a small subset of the tests are relevant, given the area you are changing, yet there’s no easy way to know exactly what tests are relevant.

Launchable’s ML engine learns this by studying the past changes and their test results. Raw information from Git repositories and test results from CI systems are refined into more meaningful data, then they are used to train a engine. The refining process also removes sensitive information, so your security people are happy.

The resulting prediction can be used in so many ways, depending on where in your software development cycle you deploy Launchable:

  • Intelligent Integration Tests: Your integration tests take 3 hours to run, so you only run this nightly. Use Launchable, and create adaptive 10% subset of it, and run this every 30 mins. Now 50% of the regressions get caught right away, while the context is fresh in your mind, instead of discovering it a day later.

  • Pull Request Validation: Every pull request you create has to first go through one hour CI validation test. Then your colleague comes in for code review, which results in a small documentation change. You have to wait for a whole hour again for CI validation test before you can merge the PR and move on. Use Launchable, and get notified as soon as you hit 50% confidence level, so that code review can begin 5 minutes into a CI validation, not an hour.

  • Local Development Loop: You are busy coding the next feature into your IDE. Code all compiles now. You know you are not done yet, but you don’t know what’s missing. So now you run tests to see which ones you broke, and you know what needs tweaking next. No more full test run on your laptop while you are having a lunch. Use Launchable, select adaptive 1%, and get 10 test failures right away. Keep tests running as you start tinkering source code, to get more feedback.

%d bloggers like this: