Disney Global SRE - Delivering Shared Service Magic

All right. So the next speaker is Jason Cox, director of platforms and SRE at the Walt Disney Company, co-presenting with Amy McCain, senior manager of SRE. For those of you who don't know, I met Jason a decade ago, and I remember so vividly when I finally got a chance to follow him around for a day.

It was 2014 on the Glendale campus. We were at the Starbucks at about 8:00 AM, and he was walking me through his calendar, describing what his day was going to look like. While he does this, a woman in the room sees him, runs over to him and says, thank you, thank you, thank you so much, Jason, for saving our team two weeks ago.

He later tells me that she was the CTO of one of the business units.

That day I got to see many amazing things. But the most amazing thing was that I got to see that scene of him being thanked by people who see him over and over again across the Disney enterprise. I've joked in the past that I got to see him thanked more times in that day than most ops practitioners get thanked during their whole careers.

Jason spoke at the very first DevOps Enterprise conference in 2014, and without a doubt, his work has influenced so many organizations, helping them rethink how their ops organizations can create better value and be more valuable to those who depend upon them. One of them was Fernando Cornago, now VP of Digital at Adidas. I asked Jason if he could share the continuation of the story and share a meaningful testimonial from someone about why their work mattered to them.

So here to share their journey is Jason and Amy.

Oh, that's wonderful. I love that. I love that we actually have that in one of our offices — it's written on the office there in Orlando: hi ho, hi ho, it's off to work we go. Well, here we are. I'm so happy to be here, and more happy that I get to have Amy McCain join me to tell you a little bit about our story around shared service magic at Disney.

100 years ago, Walt Disney had a dream. He dreamed of a company that would inspire and entertain people all over the world through the art of storytelling. For that dream to come true, it required the most important thing on this planet: people. Artists, engineers, imagineers, cast members, and most importantly, all of you, our guests. As we reach our hundredth year, the company has put together a video to celebrate that dream. Watch this.

Open your eyes. Think of the one place you've always wanted to see. This is where our magic comes from.

They say, if you dream a thing more than once, it's sure to come true. Everything you see exists together in the great circle of life. There are stories about what happened — it's true, all life.

In every job that must be done, there is an element of fun. Life is full of possibilities. What do you want to be remembered for?

The world is in our hands and we've got to do something with it. I just want to say a word of thanks to all the artists, the workers, and all of you, the young and the young at heart. You help make this dream come true.

What do you think? I've got to admit, I've seen that like dozens of times, and every time I choke up. It's what we do, right? It's what we do at Disney. So I want to tell you a little about how what Amy and I do helps deliver on that dream. And that's through our shared services team, global SRE inside the company.

So this is where we are — we're part of our corporate enterprise technology team. We support all these different amazing brands that are creating the magic across the enterprise and for our guests. But you see, there's a problem with shared services, centralized services.

Let me ask, just to poll: how many of you, by raising your hand, would say you're part of a shared services team? Oh, there's quite a few of us, more than I think. And let me ask you this second question: how many of you have to use a shared services team in your organization?

How's that going for you?

I get it, right? That's the problem. Why is it like that? Well, I think it starts a little bit like this. A shared services organization, a central corporate team, comes into the business and says, hey, we're here to help. But unfortunately what happens is the business says, no, you actually don't help. And boy, is that true. Or better yet, as Grogu would say — well, we're not really sure what he would say, but after all, he is just cute.

All right, so what's a model that works? We were looking for a model that works. How can we deliver compelling service to our businesses that actually matters, that they actually want? Well, here's what we deployed. We started embedding SRE into all the different business units, all the different product teams that needed them, all across the enterprise.

Now, I know what you're thinking. When I showed this slide to our partners at Lucasfilm, the CTO there commented and said, hey, you know, Jason, that sure looks like a Death Star. That was not the intention there. But that wasn't the only thing he said. He also said, but your team is not like other shared services teams. Your team actually understands our business. And we kept hearing that same message over and over again from different business leaders inside the company.

In fact, I've invited one of them to come here today. Nick Cannon — no, not the comedian — the head of the technology team for Walt Disney Animation Studios. I've invited him to come and speak to you about his perspective, first of all about his team, but about shared services and what he's looking for. So with that, over to you, Nick.

Thank you, Jason. And hi everyone. I'm so sorry I can't be with you in person at the summit today.

As CTO at Walt Disney Animation Studios, I'm very fortunate to be responsible for the technical capabilities of our feature animation studio, the heart of Disney, and where it all began 100 years ago. We're a close-knit group of around 1,300 people delivering long-form animated stories. Our focus is quality rather than quantity.

We tend to deliver only one or two productions a year, and each one typically runs for around five years from the original inception of the idea all the way through to final post-production. And we do it all in-house in our studio.

In animation, technology is the medium — from the very first animated feature film in 1938, when Snow White pioneered the use of the multiplane camera, through to our upcoming feature film Wish, which is pioneering a new visual style blending the timeless visuals from our earliest feature films with the latest in computer graphics technology. We have to keep the pipeline of innovation flowing to continue to elevate the artist experience so they can more fully realize their potential to deliver ever more ambitious and impactful stories.

To do that, we have a studio technology team of over 150 people working with each production crew, each of which consists of up to 500 amazingly talented artists, production management, and technicians who make the movie magic happen. The shows can't do it alone, though, and that's where our studio technology team comes in, providing them the infrastructure, services, and tools that are essential ingredients to creating our stories.

One of the things I love about what we do is the sheer breadth of technology we get to work with. On top of the regular IT responsibilities, we maintain high-performance compute and storage clusters, digital projection systems, music and video editing services, and software to do everything from crew scheduling to complex digital asset management, to physically accurate simulation of how light interacts with materials. And that's just scratching the surface.

It's a lot, and we can't be experts in it all. So we have to focus on the areas that deliver the most value. Our technology team has deep expertise in computer graphics, high-performance compute systems, and understanding how our studio functions to create our stories.

That is where the most value lies for us. For other areas, we need help, and that's where Jason and his SRE team, and other teams like them, come in.

So on one hand, we're providing shared services internally to our productions, our shows; on the other, we are consuming shared services from our friends in enterprise technology. At this point, I have to confess that I'm not a fan of the word service, and it's not a word I typically use to describe how we work. Service implies a power imbalance.

I actually looked up the definition, and it said that service is the action of helping or doing work for someone. And that's not really how we should be working across functional boundaries. I expect people to do work with someone, not for them.

So we talk about partnership, not service. It's about give and take, not fulfilling requests.

What's optimal for one team tends to be in tension with what's right for another. Managing those tensions and keeping them in a healthy balance over time is essential to our success, and is a large part of my role as a technology leader. That balance comes through partnership, respect, and constant diligence and effort.

When I first met Jason many years ago, I found a like-minded leader who wanted to learn from us as much as help us out. It didn't feel like the more traditional engagement, which I'm sure many of us are familiar with — when a leader comes in with a little too much self-belief, not quite enough curiosity, and leaves the impression that their primary interest is to persuade us to conform to the company standard. That traditional approach tends towards a lowest-common-denominator service for everyone, which diminishes the value of what is special about each corner of the organization.

It simply doesn't work in an organization like Disney, which thrives on creativity and quality.

Shared services teams need to support and amplify the value of the teams they partner with. What do I want from a shared services team? I want partners.

I want them to bring a valuable perspective, expertise, and experience that we don't have, and a curiosity and acknowledgement that they don't know it all. They should also respect the work that we do, just like we should respect what they do. I'm so thankful that this is what I found in Jason and Disney's global SRE team.

And that is the approach that I expect our technology teams to take in the way that we work with our internal production teams. Thank you all for giving me a chance to share my perspective today, and thank you, Jason, for your stellar partnership.

Ah, thank you, Nick. Did you catch some of that? Wasn't that amazing? But look what he said — I love this part. I want partners; shared services teams need to support and amplify the value of the teams they partner with. Isn't that gold? That's powerful. If you're a shared services organization, take notes. We did.

But we heard not only from Nick, who by the way has been a great partner for us throughout this entire journey, but many others who said our team was valuable, that we do a great job for them. They thanked us. They said we're a fantastic team, that we help deliver world-class outcomes.

They appreciate our innovation, our creativity. They say we have a great attitude — which I'm glad to hear went in the performance review. And they thanked us for being an awesome partner, just like what Nick was talking about.

Well, that gets to the lessons we learned. The first thing was, we needed to be curious. We need to listen — not show up to tell, but show up to listen.

Isn't that right, Amy? The number one complaint from the businesses we work with about the shared services org is that the central team doesn't understand their objectives, their challenges, their needs, and their pains the way that they do. They want someone to come in and listen, and to hear how we can help.

So the first thing we do is find out how each business ships the magic and what it takes. We need to know the mission, build those relationships, gain that trust, know the teams, and become part of the team by embedding. So lesson number one is: listen, know the business, know the mission, know the team.

The second is related to it. We heard from Nick — he said, I don't want you to work for me, I want you to work with me. That requires a change in the frame of reference: understanding their frame of reference, moving to the other side of the table. Understand what the business really needs by having empathy, so that we don't come in like the Star Destroyer coming in to do something for you, or maybe to you. Let's not do that. Let's be the fellow family partner coming in to do work with them.

What's not incredible is something like this, which I'm sure you've all heard from a central services org: you need help? Oh, submit a ticket. Instead, more like this: proximity-powered empathy engineering.

It sounds like a great product, right? But seriously, what does that mean? Go to them.

You want to see, you want to understand — go to them. The empathy comes by the geography; that is, be there, on location, wherever they are. That's a picture of our embedded SREs.

Wherever we are building anything new, we want to be part of that, understanding the challenges and the frustrations, helping work through those together as a partner. We help build empathy by being there, literally working alongside people. That shares the mission, shares the struggles, and that lets us share the wins.

All right, so if you're writing these down, here's the third one. Try this one: actually help. I know — mind-blowing. It actually makes a difference. And I love this quote by Taiichi Ohno, who says, do something to help them; if you do, people will expect that you can help them and they will look forward to seeing you again. And I love to reverse that and say this: if they don't look forward to seeing you — in fact, if they avoid you — you may not be helping. You might want to revisit that.

So this is what we did learn: how could we help? They want help creating and shipping content, products, and experiences better, faster, safer, and happier. And that's something that Amy does all the time with her team. So walk us through some of those things.

Okay. So one of the main things we do when we come in as SRE is seek to increase reliability levels and help deliver better products and experiences, both for our guests and our cast — just raising everybody up. And we want to help remove friction and reduce interaction with those unnecessary processes and procedures that don't actually help get things done and add value.

So we try to automate everything. Automate — join the rebel alliance. But we also have to be safer.

This is super important. And so we help shift left: automated governance. There might be a great book on that — talk to Topo if you need any help with that.

So security by design, super important. And happier: we've been hearing it at this conference quite a bit. How do we improve the experience for engineers?

DevX, they say, developer experience. I would like to propose OpsX now — not the budget line item, but operations experience. Development all the way to operations should be a happy experience.

Does anybody agree with me?

So what we want to do is remove the frustration and toil. Let's make sure that we have a humane operating environment for our teams and the culture. All that means this: that we can hand back time to the engineers to do more creative things. Our artists have more time, empowered to do the next thing, which amplifies the creativity across the company. And let's just focus on outcomes.

Speaking of outcomes, talk about winning. Great example: we were recently brought in to help out with a new team. The mission was to get some, let's say, slightly neglected legacy infrastructure capable of running a critical app, in time to have it stable and available for shipping the magic — which in this case was all 600-plus versions of Guardians of the Galaxy 3 that went out.

So we brought in a few of our top SREs, some of whom had worked with the technology, some of whom had worked with the people, just to build that trust right out of the gate, to troubleshoot alongside them. But it wasn't working. This stuff is old.

We had failure after failure. We had a lot of knowledge gaps. But listening to the business made it increasingly clear that our actual mission was just to get this app up and running.

So, quick pivot: brand new, shiny Kubernetes stack, lightning-fast migration, and in a few hours the app was back up and running. It was such a privilege to work alongside those teams to deliver fun stuff like this. And it was nice to see it appreciated by some of our important partners.

Actually help — no tickets, no project plans, just get it done. Build trust, build community, build the magic together.

So there you have it. That's the list: 1, 2, 3. Listen, have empathy, and actually help. And speaking of help, what are we missing? Maybe ChatGPT — a little bit of gen AI. And then what about shared service examples, or SRE, things like that?

Well, I'm going to end here, but I love this — many of you have seen this. If you've seen any of my presentations, I love this quote. This applies not just to Disney, but to this entire community. There's really no secret about our approach. We keep moving forward, opening up new doors and doing new things, because we're curious — and curiosity keeps leading us down new paths.