Kedro Meetup Starting Soon

This was our fourth session on ML monitoring and the final did not fail to go out with a bang! Amy from Fiddler did a deep dive into 3 kinds of model drift that you can encounter while models are out in the wild. I loved her description of each of these as they pertain to a loan application model.

Concept Drift - An income level that was earlier considered creditworthy is now considered riskier.
Label Drift - A larger proportion of credit-worthy applications start showing up.
Feature Drift - Incomes of most applicants increase or decrease. Or you suddenly get more applications from one region.

*This was a sponsored meetup. Huge thanks to the Fiddler folks for their continued support of this community.

Chris Albon has become a bit of a legend over the years. If you are anything like me you have found immense value in his website where he documents some key learnings he has had during his time in the ML space.

This Coffee Session Chris did not disappoint. He was fully transparent in the way that he has been running Wikimedia ML and the current constraints they have. From everything needing to be open source to on-prem he walked us through what kind of tools he wishes existed for those who are in his precarious situation.

Oh and by the way, we got another guest host to join me this session! Neal Lathia aka the modelstore guy aka Monzo ML Guy came to bring his insights and ask questions of Chris along with me. Tune in wherever podcasts are found or click the button below to watch the video.

In the beginning of 2010s the software industry came up with the 12-Factor App methodology which had a tremendous impact on the way people build software so it can be deployed and operated easily.

That change in culture combined with the later introduction of Docker was a boon for advancing the DevOps maturity of teams around the globe. Ten years later the data applications world is struggling with getting pipelines to production due to problems that can be easily solved by a similar cultural shift.

This talk will show how Kedro can help you build data apps in the spirit of the 12-Factor App methodology and advance the MLOps maturity of your teams.

Sub to our public calendar and click the button below to join the meetup. Jump in and learn more about ML Drift!

Happening very very soon! come join us!

A few weeks ago I asked in slack if we should start doing AMAs with different people in the MLOps space. There was a resounding yes from you all.

Artem Yushkovsky has taken the lead on this one and elected me (Demetrios) to be the first guest on this new adventure. I was a bit hesitant cause I'm not sure if anyone wants to hear more from me!

This Thursday from 6pm GMT / 10am PT I'll be in the newly formed community slack channel #ask-me-anything to do just that, answer questions.

Since this is a new initiative and I am not sure how its gonna be or if anyone will show up to ask questions it would mean the world to me if you came and asked something. Otherwise we can put it in the failing fabulously bucket for the year 😆

We plan on doing these every month and I promise the next one will be 🔥 I cant say who yet, but we are in conversation to get a few engineers from a little music streaming company you might have heard of to be our next guests.

Sub to our public calendar to stay up to date or jump in the slack channel to be a part of it!