Espresso and Community Service
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Background: Last week I asked you to send in a short video talking introducing yourself and what you like about the community. So far I have received a whopping total of 2 people that sent me videos!
Dont overthink it, just send me a quick video of what ever you can say! Here is an example of mine.
|
|
|
|
|
|
|
|
|
|
|
|
|
The How: If you weren't able to make it last week to our kick off meetup to talk about the Engineering Labs here is the quick and dirty on what we talked about. The goal of this project is to get hands-on experience with tools we wouldn't otherwise be using. A way for us to grow and work together with
other members of the community.
The Constraints: For the first voyage we will be grouped into teams of 3 or 4 people and we will be creating using MLflow with Pytorch. You can check out the who rundown of the slides we talked through last week here.
Get in the slack channel! Want to join the party? It's all happening in the slack channel dedicated to the Engineering labs titled yep you guessed it.... #engineering-labs. Today is the last day to fill out this form to be included in the first voyage. from here we will be forming teams and starting to find problems we can solve within the constraints of MLflow and pytorch.
More Good News: This is the first iteration of the engineering labs, we hope to quickly let this grow into much more than just an mlflow + pytorch pipeline and are very open to having you create more than that with your team! Feel free to reach out to myself or @Ilnardo92 if you have any questions around this.
|
|
|
|
|
|
|
|
|
|
|
|
|
In episode number 5 of our mini-series around Data Privacy, we dive into the term which I am sure many of you are already familiar with; DataOps.
With the dissemination of data, there’s a need to have proper processes in place to be able to follow the data and above all to make it easily available. So how does DataOps relate to privacy?
Special Guest: For this week’s episode, we have brought on Lars Albertsson, founder of Scling and former Spotify Data Engineer, to tell us all about DataOps. A chat about the right processes to use and protect the data while digging into the best that Big Data has brought us. Fabiana and Lars also examine where there is overlap when it comes to MLOps.
The is one of my favorite episodes of the season, as since data security is cropping up more and more in conversations I feel its also quite pertinent
This series is brought to you by YData. YData offers a dataset experimentation platform with synthetic data generation that makes the process of building datasets
take a fraction of the time and cost that they used to.
|
|
|
|
|
|
|
|
|
|
|
|
|
The Lowdown: We had a double shot of coffee sessions this week with two amazing guests! First up was none other than community member and head of ML at Monzo bank Neal Lathia. And on Saturday we talked to Benjamin Rojogan.
Neal, David and I had a deep dive session on Monzo Bank's infra and how they are doing ML. We also touched on some of his greatest learnings over the years while trying to put ML into prod. One of my favorite
highlights came when we touched on how they go about assessing if a problem even needs ML in the first place.
We also dove into what the structure of the Monzo ML teams look like and how they manage to efficiently collaborate. Consider this the Monzo case study you didn't know you needed. Watch here listen here.
Double Shot
Speaking of collaboration, our second shot of espresso was Mr. Seattle Data Guy himself who came on and talked to us about how he manages to stay so prolific writing, learning and working. We also dove into collaboration real deep! Some highlights from this conversation with Benjamin were around how to stay accountable while working on ML projects, the power of proper communication
and my new pineapple shirt. Have a listen in podcast land here or click the button below to watch.
|
|
|
|
|
|
|
|
|
|
|
|
|
Yep that is right we've got another member of the Netflix Metaflow team on the meetup this week! Savin Goyal is coming on to break down what learnings he and his team have had since open-sourcing the MLOps tool Metaflow.
Plan to get a good dose of the challenges encountered and choices made by the Netflix ML Infrastructure team while developing tooling for data scientists. He already told me I can't ask any questions around the next season of Stranger Things, but I'm going to anyway. (I'm such a rebel). Bring your popcorn!
Side Note: if you haven't read some of the stuff that the Netflix team is
writing about ML I highly encourage you to do so. I'll just leave this here.
|
|
|
|
|
|
|
|
|
|
|
Backstory: I feel like community member Phil Winder has been writing his book about reinforcement learning for our community! It's been a long time coming since he shared some of the unedited chapters with the community last summer. Well, the wait is over! The book is finally out for us to enjoy yeeeeehaaaa!
The Raffle
Phil was a great sport when I asked him if he would donate a few free copies to the community, so this week we are going to be raffling off 2 copies of his new book in slack. All you need to do is fill out this form and its off to the races!
If you are not one for raffles you can also purchase the book on Amazon via the link below.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|