June '21 Heartbeat

This month you will find: πŸ—Ί Navigating the MLOps Landscape, 🧐 Our MLOps philosophy πŸ“– MLOps learning opportunities, πŸ’» R with DVC, πŸŽ₯ Conference videos from our team members, πŸš€ Info on our growing team, and more!

  • Jeny De Figueiredo
  • June 18, 2021 β€’ 5 min read
Hero Picture

From the Community

This month I'm going to take you on a thought provoking journey through some of the content from our community.

So many choices...

LJ Miranda's Triad of order

The MLOps tool landscape can be confusing to say the least.
LJ Miranda, in a well written three-part series lays out a framework for making sense of this space. The list of tools is not exhaustive, but the framework and thought process for evaluating the tools is intriguing. Additionally he encourages thinking about the skillset of the members of your team within this framework to help you make decisions on the right tools. It's not just about the tools, it's about the people!

As you can see DVC makes it into the "Trial" loop, but we think we will be be making it into the adoption region in relatively short order. πŸ˜‰πŸš€

LJMiranda Making sense of the MLOps Landscape

Found in the MLOps Community

You can find more comments from LJ Miranda and others in response to a great question from AndrΓ© Godinho in the MLOps Community Slack (see below). If you're into MLOps and you're NOT a part of this Community, you should be. You can join their Slack here.

I have recently came across with DVC by listening to MLOps Coffee Sessions #6 with David Aponte and Elle O'Brien (Such an interesting talk! πŸ’―). This tool integrates smoothly with Git, tracks models & datasets, and also has an online UI DVC Studio πŸš€. Is there any use case of MLflow that DVC can't handle? I find DVC to give more rise to creativity as it integrates really well with Git. - AndrΓ© Godinho

Neda Sultova's Tutorial and Tool Rubric

Drilling down to the next level, I give you this tutorial by Neda Sultova. Not only is it a great tutorial of DVC in and of itself, but Neda also defines a clear framework for the decision making process at Helmholtz AI. Among the needs are reproducibility, workflow integration, exchangeable backend, framework agnostic, open source, and the ability of the solution to be tweaked to the team's needs.

Exploring DVC for Machine Learning Pipelines in Research (Part 1)

The first of a multi-part series on the search and decision making process for MLOps tools at Helmholtz AI.
Exploring DVC for Machine Learning Pipelines in Research (Part 1)

Our Philosophy

And at last I bring you to "The Road to AI Hell Starts with Good MLOps Intentions" by our CEO Dmitry Petrov which explains our philosophy in the MLOps space. You will learn about the experiences that led to developing our tools, what we think is the right way to solve MLOps challenges, and how we do it.

Teams made up of data scientists and developers should be able to define their own workflow based on their business requirements and team preferences, just like they do today when constructing any other software artifact. Rather than a platform forcing teams to embrace a highly opinionated workflow, they can employ flexible tools such Git, GitHub, and their existing CI tools as they see fit. - Dmitry Petrov

The Road to AI Hell Starts with Good MLOps Intentions

Dmitry Petrov explains the journey and philosophy at the heart of Iterative.ai's MLOps tools.
The Road to AI Hell Starts with Good MLOps Intentions

Big News! πŸš€πŸš€πŸš€

In case you missed it, June 3rd we introduced our latest tool: DVC Studio! A web application that GUI display your team's work with DVC and CML. We know this has been on our Community's wishlist and now it's here! You can check out all its features and give it a try here. Or check out the introduction video below.

By clicking play, you agree to YouTube's Privacy Policy and Terms of Service

Learning Opportunities

R for DVC!

Are you or someone on your team an R user? JoΓ£o Santiago who has contributed to DVC, recently came up with "dvcru" to provide utility functions for DVC pipelines using R scripts. Additionally the project aims to show typical workflows they enable as well as provide project templates. Check out all the R goodness in this Github Repository.


JoΓ£o Santiago's repository for dvcru, providing utility functions for DVC Pipelines using R scripts.

Milecia McGregor at PyData SoCal

Next up we have Milecia McGregor presenting and live coding at PyData SoCal organized by Pramit Choudhary. Check out her talk on "Reproducible ML Experiments (with Git and DVC)" and all the great questions that ensued.

By clicking play, you agree to YouTube's Privacy Policy and Terms of Service

Dmitry Petrov at MLOps World

Finally we have Dmitry Petrov's talk at the MLOps World Conference about machine learning in production entitled "Data Versioning and ML Experiments on Top of Git."

By clicking play, you agree to YouTube's Privacy Policy and Terms of Service

DVC News

We're still growing! Meet this month's new team members.

New Team Members

Jelle Bouwman joins us from Utrecht, Netherlands as a software engineer. He's worked as a consultant and at an agency. He's most proud of the work he did with his team at the Port of Rotterdam. In his free time, Jelle loves reading fiction and books on human psychology/productivity, hiking and making music with others. He has already shared with the team a great playlist to listen to while trying to focus! Welcome Jelle! 🎼

Next we welcome Alexander Gushcin. Alexander joins us from Russia where he has been a Data Scientist/ML Engineer for the last five years. He's also participated in many Kaggle competitions and was ranked 5th in general competitions at some point! This led him to create a Coursera course on how to win data science competitions about the tips and tricks needed to win one. Teaching is his passion and you will probably see him producing some content in the near future. πŸ§‘πŸ½β€πŸ’»

Mikhail Sveshnikov also joins us from Russia where he formerly worked as a Data Engineer Team Lead for Rubbles. He created ebonite, an ML deployment tool and teaches Python and Big Data at HSE University. Finally he is one of the admins of ods.ai community, which creates global projects to unite the community, promote Data Science, and help people develop their skills. In his spare time he likes to play guitar, badminton, ski, and mix cocktails. 🍸 Cheers Mikhail!

Jervis Hui is joining the go-to-market team at Iterative and is from NYC. He's worked in product marketing at various Silicon Valley tech companies over the years and is excited to bring his experience to the open source world of Iterative. He's passionate about D&I in hiring and looks forward to learning from everyone! We're excited to have Jervis on board! πŸŽ‰

Open Positions

And yes indeed, we are still hiring! Use this link to find details of all the positions including:

  • Senior Front-End Engineer (TypeScript, Node, React)
  • Senior Software Engineer (ML, Dev Tools, Python)
  • Senior Software Engineer (ML, Data Infra, GoLang)
  • Machine Learning Engineer/Field Data Scientist
  • Developer Advocate (ML)
  • Director/VP of Engineering (ML, DevTools)
  • Director/VP of Product (ML, Data Infra, SaaS)
  • Director/VP of Operations/Chief of Staff

Please pass this info on to anyone you know that may fit the bill. We look forward to new team members! πŸŽ‰

Next Meetup

Don't miss our Meetup June 24th at 3:00 pm UTC (8:00 am PDT), where Sami Jawhar of Kernel will present different experiment use cases. Bring your questions and thinking cap! It's bound to be a great session!


June DVC Office Hours with Sami Jawhar of Kernel presenting experiment use cases.

Tweet Love ❀️

Do you have any use case questions or need support? Join us in Discord!

Head to the DVC Forum to discuss your ideas and best practices.

← Back to blog