You have to bring AI tools into your CI/CD workflow. We've got it covered. Watch this

The Register - Fri Sep 24 13:03

Special series After a great first episode, the MCubed webcast will be back on October 7, 2021 to tackle a whole other beast: Continuous Delivery in Machine Learning.

While you might be familiar with the intricacies of Puppet, Ansible, and co to speed up and secure your project delivery, transferring that knowledge into the ML operations space isn’t exactly easy. After all, it’s not just application code and maybe config files you have to be aware of anymore. There are more disciplines involved than usual, which means you also need to consider training data, models, and the tools and processes needed to create such artifacts when setting up (at least partly) automated processes.

Luckily, you’re not the first person who has to try and get a grip on this complexity. For the second installment of our machine learning practitioners webcast, DevOps and data expert Danilo Sato has been kind enough to agree to walk you through the components needed in a good continuous delivery system for ML applications.

Having worked on all sorts of platform modernization and data engineering projects, Sato has built up a wide knowledge of what works and what doesn’t. He even shared his findings in the 2014 book DevOps in Practice: Reliable and Automated Software Delivery. Since then, he was able to dig deeper into good practices for data science and machine learning projects, which has turned him into a well-rounded discussion partner on high-quality continuous delivery in complex setups.

During his talks, Danilo will help you understand what is necessary to provide users with a stable ML application that makes use of all the latest assets. He’ll make you aware of tricky bits, such as quality assurance and testing, so that you don’t startle at first contact, and point out things to keep in mind when setting up workflows for reproducible and reliable releases.

The webcasts will kick-off at 11am BST (12 noon CEST) with a quick roundup of the latest machine learning-related news for software developers before diving into the world of continuous delivery. If you register, we’ll even remind you on the day in case you forget the time over an especially tricky problem.

Speaking of which: the MCubed webcast is meant for practitioners like yourself. So if you have any topics you’d like to learn more about, want to share some useful tidbits yourself, or are interested in practical experience reports from certain industries, reach out and let us know so we’re able to incorporate it into an upcoming episode.

We’re looking forward to seeing you on October 7! Don't forget to sign up here.