Host Intro (Gene Kim)

Thank you, Girija, Danai, and Jennifer.

So I met the next speaker, Dr. Angel Diaz, almost seven years ago, back when he was in the IBM Research division as the executive in charge for advocating for open source, developers, and cloud.

I was delighted to reconnect with him in his new role as VP of Technology Capabilities and Innovation at Discover Financial. I learned so much about Discover: that it goes far beyond just being the third-largest card brand. If you look closely, you'll notice that they are what powers Apple Pay, which I absolutely love using. And they have over 8,000 developers.

Dr. Diaz is responsible for helping elevate developer productivity and innovation and helping the entire organization learn. He will be presenting with Sheila Lodhia, Vice President of Business Technology, Application Development for Card.

Here's Angel and Sheila.

Dr. Angel Luis Diaz

Hello, everyone. I'm Angel Diaz, Discover's Vice President of Technology Capability Innovation, here with Sheila Lodhia, a colleague, friend, partner, Vice President of Application Development.

We are living in a technology-fueled business renaissance. The art of the possible is only limited by our imaginations. So the question is, how do we wield that power responsibly? And that's what Sheila and I would like to talk to you a little bit about today: how we are building our craft with Discover to continue our digital transformation and help our clients achieve a better financial future.

With that, let me hand it over to Sheila just to tell you all a little bit more about Discover.

Sheila Lodhia

Thank you, Angel. So this is a perfect slide of who we are today and where we're focused on.

Discover started out as a credit card company back in 1986, accepted only in North America. And today, we're accepted in over 200 countries and expanded our credit offerings well beyond a credit card. We have home loans, personal loans, student loans, and a variety of deposit products. We started out with no physical presence and stayed on course to be today's leading digital bank.

And to continue that leadership, we actively strengthen our foundation by growing internal capabilities, especially in spaces like engineering, data analytics, and product ownership. And for all of this to come to life, it is all being done under our Runway Program, a program where both business and technology come together and improve on that product accountability, automation, and simplicity across the company.

This is taking Discover to the next level of agility and ownership. So let's take a deeper look as to how we're accomplishing that.

Dr. Angel Luis Diaz

Thanks, Sheila. We're going to start with getting a little bit deeper on how we're creating a runway for digital transformation, specifically how we're building our craft with something we call the Discover Technology Academy Community. That's about building our products, building our communities, building our career, and continuously building our innovation. And we'll close it out with a little fun thing about how we're building our house together.

We have had leaps in technology that have then been paired by democratization of who can deliver value and how quickly you can develop value. However, this time it's different because it's not just about technology. It's about people, it's about process, it's about how you organize yourself. It's about communities and ecosystems. So it's all that paired together, and that's exactly what we're trying to accomplish here with the Runway Program.

We're going value stream by value stream across card, bank, payments, thinking about how, centered around our customers, how we can better organize ourselves to meet those demands, to organize our teams, to take those teams and then organize them in a way that is isomorphic to our architectures, to our code patterns, to our infrastructure, and becoming a product organization and a platform organization. That's at the heart of our Runway Program.

It started in 2020. We're kind of ramping up, seeing impact today, and then next year we start to fly.

There are many ways in how we measure our success and our ability to improve with our Runway transformation. Our ultimate objective is really to provide our customers a brighter financial future. So how do we do that? Well, we've got a series of OKRs that go down from business results, customer engagement, all the way through to what we do in engineering.

We look at engagement of the communities, number of teams that go through dojos, the improvement of those teams over time, whether it's velocity, burndown, story points committed, story points delivered, our DORA metrics, our uptime, our availability. All of those things are metrics that we use as a compass to make sure that we're headed in the right direction and we're headed fast enough to achieve those objectives.

This work that we're doing here at Discover is so important, from our executive committee, CEO, all the way through to every single engineer, business person in the company. Let me hand it over to Sheila to talk a little about how this comes alive in practice. Sheila, over to you.

Sheila Lodhia

Yeah. Okay, so now we have our product structure and our product team. How do we make sure all of these teams are applying the Agile and DevOps concepts consistently?

One way is through DTA's Dojo experience. A dojo is truly a hands-on learning experience with a coach, where this coach is sitting side by side with the engineers and the product owners on building their DevOps capability. And the best part is they're applying it to the actual work from their product's backlog.

So let's take a look at DTA's offering of a Dojo experience.

The Discover Dojo is a four-week immersive learning and holistic transformation model where engineers pair together while still doing work with the objective of improving their way of working based on well-defined metrics. Now, as teams go through a Dojo, they not only leverage the experience of our coaches, but also actively contribute back and improve the craft so that others can share within the Discover Technology Academy experience.

It's as simple as taking a piece of documentation, doing a pull request on GitHub, and then adding your knowledge to the system. Every product team has a home in Discover. Their craft is continuously improved and shared and reused across Discover itself.

So now the question is: are we getting faster and better as we apply these DevOps practices? Well, that is where we're using the DORA metrics to help us measure: hey, how fast are we getting the features out to the customer? How predictable is our delivery? The maturity of our engineering practices in terms of releasing changes into production. What's the quality of those releases? And how fast can we restore back upon a failure?

To actually see the results of these Dojos and DORA metrics, we selected three product teams that recently went through a Dojo. Now, all of these product teams are modernizing their entire stack from legacy platforms to a cloud-native app, providing many opportunities to reap the benefits of DevOps.

All right. You're about to meet these three product teams. They are key to Discover in terms of customer experience and key financial metrics such as sales and receivable. You are going to see how they're in the beginning stages of their DevOps journey. But at the end, we should see the team become a lot more autonomous, because as of right now, they have multiple dependencies, especially because how their architecture is, well, one big ball of spaghetti code with very high technical debt.

With them achieving team autonomy, they can drive the product vision end-to-end independently and release it to production six times faster than today. And I'm saying this based on my experience. It actually unlocks a whole new world for the engineers. They are now at the table driving that product vision with that product owner, and no longer in a closet churning out code with no idea how it provides value to the customer or the company.

Meet Team CAD. They are a platform team where they are decomposing 16 monolithic SOAP services that have about 40 methods to them each. They support the actual entire card business. So these are critical services. And we're converting them to individual microservices on an event-driven architecture. And as each component is stood up, it will be federated back to the product team that owns the data.

Let's meet the next team.

Team LIR, where they are modernizing the entire customer journey that allows customers to request a credit line increase. This transformation is taking a legacy mainframe WebSphere application to a cloud-native app. So these teams are learning core Agile essentials, TDD, blue-green deployments, and testing automation.

And finally, our last team, the Authorized User team, where they're also modernizing a legacy mainframe WebSphere app to a cloud-native app. This customer journey supports the ability for a customer to add an authorized user to their credit card, and it is very key to driving sales for Discover. This team is learning also CI/CD with blue-green deployments and canary releases, TDD, test automation, and UX optimization.

So let's hear from all three of them on their Dojo experience.

Team CAD Representative

So basically my team is called Customer and Account Data team, and it's a platform team within Discover, within portfolio value stream. So our goal in this team was to break apart these monolithic legacy aggregators into domain-centric APIs.

When we started in this team, I had three employee engineers, and they were not exposed to the new Agile ways of working. And before Dojo started us, we pretty much were on our journey to implementing the new Agile ways of working on our own. Dojo came in and they really helped us in reinforcing the concepts to the team. And what it did was it helped the team in embracing those concepts in a much faster timeframe.

And ultimately, today, my throughput for the team has increased and the quality of code is getting better and better, and we are pretty much on our path for deploying and developing and deploying iteratively.

Team LIR Representative

We are a product team. Our product is line increase request. So this product gives the ability for customers to request line increase. As a team, we were working as more of a feature delivery team, but we definitely wanted to have a refresh and went through an exercise called Dojo.

So we got seven coaches for us to help to go through this Dojo exercise. So this dojo helped us to establish a fundamental understanding on how to operate as a product team. So the results of this dojo was phenomenal. Our features were taking almost three months to deliver. Now we have six weeks feature delivery.

And also we automated a lot of things, so it's a more iterative way to kind of deliver the product itself. We have delivery that's going out that can be deployed any time, rather than waiting for a big release.

Authorized User Team Representative

So yeah, my team is responsible and owns the product called Authorized User, which gives our card members the ability to add additional users. Going into dojos, they were looking for practices they could utilize, which would help them move from the legacy system to the modernized system using DevOps as well as microservices and React, using all these at the same time, using best practices for TDD and BDD.

When we came out of this, we could see that it really helped us kickstart our process. The key successes I would say is velocity, of course. The velocity is gradually increasing, and we are looking at deploying every two sprints now. Earlier we used to do every six sprints. We are looking at deploying every two sprints starting next month.

Dr. Angel Luis Diaz

Thank you so much, Sheila, and thank you to the team. Amazing results, and it really shows our passion is shining bright here at Discover.

I want to cover two more really important elements of how we're trying to continuously improve here at Discover, and one is around community and innovation and learning.

We have this saying: when you learn, you teach; when you get, you give. We're good partners, and we develop others.

We kind of have a community, and the way we've implemented digitally and also physically is kind of what I think of as a semantic graph backed by inner open source principles. So it all starts with the individual. And everyone has the ability to kind of explain who they are, their competencies. Think of it as a radar diagram of skills. How much do I really know about Node.js or Jenkins or whatever, right?

Then you back that up with the connected to their roles in the organization, in their teams. Technology. Technology that supports that, and then of course, content: learning paths, blogs, videos, articles, patterns, et cetera.

So what does that give you? That gives you the ability to traverse anywhere in this highly connected graph. You can go from someone who's contributed a blog to a whole set of education around that topic area, to individuals who are experts in that to ask questions.

Investment in learning pays the best interest. We develop ourselves, and we always stay ahead.

So we've got different approaches to how folks can continuously learn, and we really, really do value our most important asset, which is people. We have learning paths where folks can go in and follow a set of courses, certifications, et cetera, to just improve.

We also have the ability, and I had talked about this kind of radar diagram of skills, for people to start to choose their own adventure or their learning jungle gym, right? Where they can kind of go and bring in different pieces and create their own path around learning.

Or what I find most exciting, in-context learning. You wake up in the morning, you have a particular task, you need some help. How do you go find out how to go do that or other people who know how to do that? Or maybe even surface information within your favorite IDE, right, to help you get your job done most effectively.

Now, innovation is doing new things. We at Discover, we're curious, we innovate, we simplify. But there isn't an innovation ivory tower at Discover. Innovation happens everywhere. So the question is, how do you empower that? How do you create the gym equipment, right, needed to exercise that muscle?

So we have a set of programs that we kind of have within Discover to help do that, whether it's doing an innovation design-thinking sprint within your value stream delivery, or whether it's a global hackathon or a hack-in-the-box that a team of six just wants to do an experiment. Or what we do around external communities. No company, no organization's an island. We're not an island. We're active participants in meetups, conferences, and of course, open source communities.

So at the end of the day, it takes a village. It is our house, and we build our house together. And let's just take a moment to listen to some folks and their passion around our house.

Discover Technology Academy Voices

The Discover Technology Academy experience is my house.

This is my house.

This is my house.

This is our technology house.

This is my house.

This is my house.

This is my house.

My house.

My house.

Dr. Angel Luis Diaz

Sheila and I are so proud to be part of Discover, and we're so proud to be part of this movement, so proud to be talking with all of you.

However, we can't do this alone. We're not an island. We need your help as we continue to build our house. Many different ways.

We need to collectively as an industry evolve the art of the possible. Still more that we need to do around pipelines, DevOps, et cetera, to start to really blend in not just DevSecOps, but SRE and everything else into that equation of how we deliver products.

We need, as a community, more access on standard best practices that we can easily share and contribute. Imagine this academy experience that we're talking about at industry scale where everybody shares some of their best practice.

Those are the kinds of things I think, Sheila, that are so important for us as a community and as a practice. Let me hand it over to you.

Sheila Lodhia

Great. Thank you, Angel. Really appreciate all of you taking the time to hear, like Angel said, our Discover-type experience under the Runway program and its importance in scaling product centricity and DevOps capabilities across Discover.

So please reach out to either one of us for any comments or questions, or if you just want to simply join our house. Let us know.

Till then, thank you very much.