Guide Book of the MLOps Community
Forming tech communities and getting into MLOps
--
Whenever a new technology, programming language or any other area of interest comes into light and becomes more and more popular, a new group of interested people is formed. Whether it was Blockchain, Machine Learning or even earlier — Internet of Things, it is inevitable that people will start grouping together to learn and exchange observations.
Tech People Have It So Easy
We, computer scientists or tech people in general have it easier. We live in 2021 and the Internet is the only reasonable option to build a community which spans across all countries and allows people to discuss, present their ideas and learn together. We can not only use computers and Internet with our eyes closed, we have no troubles adapting to new tools that help us communicate more efficiently — Slack and Zoom or Clubhouse are only the most recent examples.
This is why tech-related communities are formed so quickly and grow to thousands of users in a blink of an eye. Apart from communication platforms themselves, we have all that we need to learn on the Internet e.g. GitHub, e-learning platforms, conference talks on YouTube etc. No other industry has all of their resources gathered in one place like we do, at our fingertips.
Machine Learning Community
Machine Learning has been with us for a couple of years already. Of course if we think about when first machine learning concepts or statistical methods were discovered, Machine Learning is way much older than most people think it is. Older than most of us are.
According to the Timeline of machine learning, first underpinnings of Bayes’ Theorem were discovered back in 1763. With automatic differentiation discovered in 1970, RNN in 1982 and random forest along with SVM in 1995, machine learning algorithms were fully functional and available before… I was even born actually.
I don’t know if you could hear about machine learning on the streets back in 1995, but I remember when it started to gain popularity on the Internet, around 2010s I would say. The timeline I referenced before says Kaggle was launched in 2010. This is a massive community platform for data scientists and machine learning practitioners in general. This is by far the most useful resources for practicing data science in my opinion.
From 2010, there were many factors that caused Machine Learning popularity to go through the roof: better hardware, more institutions conducting ML research and publishing their findings, more people interested in tech and proving that these methods work well in business use-cases.
And one more thing we are particularly interested in: communities. If you take a step back and look at the landscape, you will notice how many options we have now to share ideas, projects and research: arXiv or Papers With Code, GitHub projects, Slack groups, Twitter, conferences, webinars, blog posts, Kaggle competitions, huh. Just to name a few.
Case of machine learning and other areas prove that whenever something interesting and promising comes up, we will form a community which together works on tools, books, projects that help the area grow.
Assessing MLOps Community
So, now that we know how big a tech community can be, how fast it can grow and how it can influence people in that group to help it expand even faster, let’s assess the MLOps Community. If you are here, I will assume you already know what MLOps is, so let me just skip the introductory part.
Beforehand, let’s look at Figure 1. Notice that this topic started gaining interest around it in the middle of 2019. At the time of writing this post: it was two years ago. If back in 2019 hardly anybody knew about MLOps, how much could have been achieved till now? How numerous and organized is this community today? Let me introduce and guide you through the list of its initiatives and on-going projects in the ML Operations community.
Slack
My #1 bookmark since I became MLOps Engineer is MLOps.community Slack, where over 4000 engineers exchange their ideas, post learning resources and answer questions related to machine learning, data science and operations.
More than that, there are channels where job offers are posted, where you can ask for a career advice or shamelessly present your recent ML project (and get feedback). And the coolest thing is that you can expect answers within hours.
PS. Yes, there is a separate channel for MLOps memes.
Weekly and Monthly Newsletters
Newsletter will give you a summary of recent activity in MLOps area, once a week, delivered right to your doorstep. Additionally, on a monthly basis you receive “MegaOps” newsletter with a more zoomed-out view on the landscape and latest trends in MLOps. This is a perfect way to stay up to date with low effort, if you don’t have enough time to go through all the papers and Slack discussions. Subscribe to newsletter here.
Meetups and Coffee Sessions
If reading is not your thing and you prefer videos and podcasts, these two series of events occur on regular basis in the MLOps community. Meetups are live every Wednesday at 5pm UK (9am PST) on Zoom but if you can’t be there, every meetup is uploaded later on as a YouTube video or Spotify podcast, just like Coffee Sessions.
Both are roughly 1-hour long talks with first-class engineers who lead ML teams in top tech companies or creators of ML tools who are involved in the community share their insights on ML & Ops concepts, tools and projects.
If you’re looking for something more concise, like short clips cut out of long discussions, check out the Meetup Clips available on the community YouTube channel.
Reading Club
MLOps Community Reading Group meets every two weeks for reading sessions. Every meeting is dedicated to one paper (or e.g. book chapter focused on ML Engineering and Operations) which is going to be discussed, everyone reads the paper beforehand and then people debate — first in small groups then as a whole group.
Exceptional initiative for people who love reading papers and discussing them with others. You can check the contents of previous sessions as well as sign up for the upcoming one here. Don’t miss it.
Engineering Labs
Engineering Labs is a place where you as a person have the opportunity to join a team and work together on solving a particular Machine Learning Engineering challenge.
As introduced in this video and described here, Engineering Labs initiative focuses on collaborative work between engineers, forming a cross-functional team to solve ML(Ops) problems together.
About the content of labs, depending on the participation and how the initiative will evolve as the community changes, we will ask you and anyone who wants to propose ML challenge to solve. We will collect them and we will let the community cast its vote on it.
And, in the end, once you and your team solved the challenge, we will ask you to share your experience and what you learn. (…)
This sounds like a great place for people who enjoy playing as a team, sharing knowledge and learning from others. Join #engineering-labs channel on the Slack to start the journey!
…and even more!
The community has much more to offer: Women of MLOps channel, informal Office Hours chats, MLOps Stacks open-source project and local meetup groups. The best place to find all of these is Slack, where you can browse all 46 channels and join whichever you like.
You will certainly find something for yourself!
All roads lead to MLOps.community
While I focused solely on “MLOps.community” and its projects, there are many other resources, blogs or YouTube channels dedicated to MLOps area. The reason I focused on this crowd is that, as far as I’m concerned, this is the biggest group and I believe everybody actively working in the area is there.
Bear in mind that the community is open for everyone and if you are hosting your own podcast, writing a blog or starting a company you are not only welcome there but encouraged to be shameless and brag about it!
Conclusion
Although the MLOps topic itself is still fresh and maturing, the community is already incredibly organized and is growing at an amazing pace. In one of the previous sections I said that it all started in 2019, but in fact MLOps group is barely one year old (MLOps Community 1 Year Anniversary Meetup).
It only took one year to organize all these projects, meetups with incredible guests and supporting community. That’s terrific! As part of this team, I keep my fingers crossed for future growth of the group and whole area itself.
