With $10 million in funding, 2 ex-Googlers launch Mabl to bring machine learning to software testing –
What do you do when your first startup is acquired by Google shortly after launch? Hang around at Google for a few years and then launch your next venture, of course.
That’s exactly what Stackdriver founders Dan Belcher and Izzy Azeri did after they sold their cloud systems monitoring business to Google back in May, 2014. The internet giant didn’t waste time in leveraging its purchase, and it went on to launch the newly branded Google Stackdriver in early 2016. This incarnation served as a monitoring and diagnostics tool for applications stored on the Google Cloud Platform and, interestingly, Amazon Web Services (AWS).
After a short limited alpha period, the company is also officially launching its Mabl QA automation product today for free as part of a public beta program. Existing clients include fitness-tracking app Runkeeper, which is now owned by Asics, and Codeship, which was acquired by CloudBees a few weeks back.
In a nutshell, Mabl is all about creating and maintaining automated tests and then analyzing the results, and promises to help developers’ quality assurance (QA) processes keep pace with the actual software development part of building a web app or website.
Indeed, QA is key to shipping functional and flaw-free products, but it can often slow shipping down because it requires continuously testing your code. Users install a Chrome extension to “train” Mabl, and Mabl creates machine learning models that predict how the app should perform, and the more tests it carries out on a particular app, the more it learns. For example, it may observe that a web page is loading more slowly than it had previously, which makes it easier to detect the flaw and determine what changes have led to this latency increase.
“We have interviewed hundreds of software teams, and their message is loud and clear: QA is struggling to keep pace with DevOps,” said Azeri. “This area hasn’t received the level of investment and innovation that we have seen in other aspects of modern development, and teams are forced to choose between quality and speed.”
There is no shortage of automated software testing tools out there, including Selenium, which Mabl said is the QA tool used most commonly used by its current alpha-stage clients. But by bringing machine learning into the mix and offering scriptless testing through the browser, the company hopes to not only expedite the testing process, but to simplify it. It can also integrate into developers’ existing workflow, as it supports Atlassian, Jira, Slack, Webhooks, Atlassian, Jira, Jenkins, and more.
“Mabl is really about making QA automation as simple as possible,” added Belcher. “You can create tests in minutes, and we handle the burden of running and maintaining them and analyzing the results.”
Mabl is available in public beta for free from today.