1063. Insights: Can traditional banking survive the AI era?

19:34Jouk Pleiter:Oh, absolutely, yes. But I also think, like, you know, guys, I think Revolut is going to destroy traditional competitors. Like, they already have the right digital art. They have also, but relatively compared to anything legacy, culturally way stronger in just bias for action and execution power. Technology stack, much more simple, because, you know, you just have, you basically can skip 30 years of legacy. And I think kind of the leadership drive to adopt AI. So guys, I personally think for everybody in the audience, we had promises of personal finance management and hyper-personalization. And we've been talking in a way, compared to what we can do today, it was all bullshit because it was within deterministic logic, you don't get far enough.

20:23Jouk Pleiter:You just don't get far enough. So I think right now we're going to see a renaissance in truly, you know, AI agents that are really almost like the digital twin of the customer. They can negotiate on them. They know their financial behavior. They can coach them, guide them, whatever legal language you need to use. But they can absolutely build a better job in going back to the intimacy and personalization you had in a little village where you know the person from the bank in person, right? It's weird, isn't it? I can get back to that loop.

20:55David Brear:Yeah, that traditional values but delivered at an amazing scale. I completely agree with that. That's an amazing opportunity, isn't it? Because the sort of reminiscent feeling of somebody really understanding you and servicing your needs in that way, that being sort of democratized to everybody now because really the person to talk to is something reserved predominantly for people of high net worth or ultra high net worth to manage your finances. But if you've got agentic capability, you can be proactively looking after people's finances for them. That's almost like private banking for the mass market, isn't it?

21:36David Brear:That's a huge step forward.

21:38Jouk Pleiter:Exactly what it is. Yeah, it is basically the white glove treatment you see in private banking, you know, at a mass scale. Of course, you need to do, you know, and of course, you know, the pre-processing, the post-processing, just make sure, like, there are still obstacles in the industry also to kind of get this thing, because the agent framework is collaborating and is truly communicating with clients. They're going to give truly consent to agents. Well, before you can get that signed off, then you really have to make sure you've got a proper architecture in place. We do, but like, I think you also do not underestimate the amount of work you need to do to kind of get risk and compliance to sign off on that scenario.

22:15David Brear:Definitely.

22:15Jouk Pleiter:I agree. It's rethinking... Half full, half empty, right? It could also be, yeah, nothing is possible, right?

22:21David Brear:So... Definitely. Rethinking all of those steps and rethinking what makes you safe and what makes you secure and all of those things. In the same way as, you know, when the changes were there for the advent of cloud technology, then it took a lot of people a lot of time to get used to that, didn't it? But all right, folks, in this half, we've explored really well, I think, why the current model is under pressure. In the second half, what we're going to do is take a bit of a look at what actually replaces it and what a AI native bank actually looks like in practice. It's all coming up after this quick break.

22:55Study and play.

22:57David Brear:Come together on a Windows 11 PC. And for a limited time, college students get the best of both worlds. Get the Unreal College Deal. Everything you need to study and play with select Windows 11 PCs. Eligible students get a year of Microsoft 365 Premium and a year of Xbox Game Pass Ultimate with a custom color Xbox wireless controller. Learn more at windows.com slash student offer. While supplies last, ends June 30th. Terms at aka.ms slash collegepc. Ready to soundtrack your summer? With Red Bull Summer All Day Play,

23:30Jouk Pleiter:you choose a playlist that fits your summer vibe the best. Are you a festival fanatic, a deep end DJ, a road dog, or a trail mixer? Just add a song to your chosen playlist and put your summer on track. Red Bull Summer All Day Play. Red Bull gives you wings. Visit redbull.com slash bright summer ahead to learn more. See you this summer.

23:55David Brear:Welcome back to Fintech Insider, where we're exploring the shift towards AI native banking. I mean, we talked a lot in the first half around where we've come from. And actually, the point that really stuck with me around that traditional values, but new digital capabilities. I mean, it feels like that sort of layering of systems and everything that banks have put together. I mean, I know we've spoke about this before, Duke, but the banking is a world of, you know, pirates of the high seas. It's mergers and acquisitions and developments of monolithic structures and all of these things. Having a system that traverses those things for you to pull them together for great experiences.

24:41David Brear:I mean, that's the dream right there, right? But I guess on the metaphor that you're drawing on, and I really like this, is It's not about point solutions from a platform perspective, but an operating system that creates a collaborative structuring. So talk to me a little bit more about why operating system is the terminology that you like.

25:05Jouk Pleiter:So if you talk to most banking executives, if you ask them, like, would you like to increase the number of systems you have or do you want to reduce them? The overwhelming answer is, can we simplify the landscape? Can we reduce them, et cetera, right? but like historically you know making we all know that we have way too many we all know we have to wait but I think historically without AI it was just too expensive and too risky to do a full transformation so you basically continue to operate in the relatively old model where every department continue to buy their own app and guys are fighting over who owns internet and who does it not and like they're all operating in their own you know either in your own channel or your own product group or you're in the front office or in the back office but I think the org structure is basically, and it's the only way to move forward, is basically giving people certain mandates to run certain parts of the total thing.

25:55Jouk Pleiter:I think that in itself, that org structure, of course, is then resulting in basically a broken operating model where you have fragmentation. And now with AI, they can all start to apply little AIs in there, but ultimately, it's not going to solve the larger picture. So we went on a mission to say, can we define a formula to help banks to kind of get out of this wrong reflex of buying more systems and ironically adding you know any system you add and any system integrator you hire you're making the problem bigger ironically and also you've got a rating where you distribute a lot of investment money into silos and then in the silos they're going to find their siloed solution so you basically have to break down that reflex a little bit like how do we budget how do we run these things etc our vision is basically hey instead let's say let's stop the number of system expansion we'll just keep what we have right and let's not replace it but let's at least now make sure that at least it starts to collaborate to fulfill the customer intent.

26:51Jouk Pleiter:So we said as Backbase, historically, you know us from digital channels, right? And that's kind of our, basically our rights to play. What we've seen there in the last 10 years is basically digital channels are becoming the primary user interface for most banks. So it's also in that primary user interface, you see all the intent. People naturally want to go there and like with voice or whatever, what they want to do, but they want to show their intent. Well, then you can see that like you can execute some of that intent fully automated with digital banking. And then there's anything else is basically exception, edge cases, which is resulting in basically, you know, hundreds if not of thousands of people that run in the bank just to basically work on these exceptions.

27:31Jouk Pleiter:So I think where we are right now is, okay, we have to move now from just digital channels into the full frontline. So let's figure out, like we started that journey five years ago, how do we make sure it's from digital channels, I can seamlessly activate people in the front office. They work on the same reality. They're part of the same workflow, the same data. I can start to find synergies between mobile and call center. Or I can find synergies between online and the relationship manager. Like this is such a natural way of bringing these things together. So that was the foundation of our view.

28:03Jouk Pleiter:Okay, we have to go from digital channels to unified frontline. And the frontline is basically the bank where you basically customers and employees are collaborating. And how do we start to kind of basically simplify and streamline workflows? And the workflow could be from new customer onboarding to new customer activation to expanding your share of wallets, onboarding them in other products and basically servicing them. And you've got risk. And then you basically arrive in, you know, 200, 400 different customer intents. Some of them are STP, straight to processed, fully automated. And the vast majority is not.

28:39Jouk Pleiter:And that's kind of, I think, then where we came in with BankingOS to say, okay, we have to bring a meta operating system that sits basically under the unified frontline. So it can steer channel interactions in mobile and online and conversational, but it can also connect it to employees in the front office. And if need be, it can escalate it deeper into operations. How do we kind of get that on one, let's say, organizational orchestration blueprint? And then on top of that, how do we start to activate agents to help customers and to help employees. Because if you're going to reinvent it anyhow front to back, then it's not just the workflow redesign, but it's also all the different tasks within the workflow.

29:20Jouk Pleiter:Historically, lots of people moving forward, a better balance between people and agents.

29:25David Brear:Yeah, it's an interesting shift that, isn't it? That sort of the experience you get is so different when the customers and the employees and the agents and the advisors are all looking at the same thing. You know, they're not different versions. They're not completely different setups. And actually, not only from an experiential perspective, but just from an operating cost perspective of that organization. If, you know, if actually, as you say, if all you do every time is add another one, then, you know, everybody's OPEX just gets bigger and bigger and bigger. And it's really the millstone around everybody's neck, isn't it?

30:00David Brear:In terms of the cost of serve of customers. But I mean, that points to a world where, you know, the shift is changing really fundamentally how banks design experiences for their customers. Because, you know, it's not frontline, you know, internal advisor designers and, you know, designers of experiences for customers being different, but them being sort of two sides of the same coin, right?

30:26Jouk Pleiter:Yeah, based on what you just said, I'm just brainstorming here. But like the most successful customers I see navigating is basically a strong alliance between the chief digital officer and the chief operating officer.

30:37David Brear:Yeah.

30:38Jouk Pleiter:Like, and of course there's the CIO, CTO stuff. But like if the chief digital officer, because then you basically own the channel, that's where the intent starts and that's where you need to do the fulfillment. If then together with the CIO, you can say cross departments, cross product groups, cross systems. Let's start to streamline, you know, one front to back workflow, which is basically from intent to full resolution and then let's figure out like how do we compress the resolution time and that's a big win for the customer because it's now instant. It's a real massive true customer experience improvement and then automatically it has a huge bonus on operational efficiency if you do it well.

31:14Jouk Pleiter:Yeah, I mean I feel like that. How to get there, I think that is really the most pragmatic way. CDO, COO and basically customer fulfillment loops figuring out like one by one, solving a massive ROI.

31:29David Brear:Well, it goes back to what you said with Conway's Law though, isn't it? If the technology can allow a different model of operation within an organization to serve much more complex outcomes, orchestrating those outcomes in a different way, it really feels like that's what unlocks the potential, doesn't it? If you look at all of the challenger banks, like the fundamentals of their service are okay. You know, like they're good, don't get me wrong. And, you know, they occasionally have a pretty innovative idea here, there or anywhere, but it's the operating model. It's the ability to make change happen.

32:06David Brear:It's the way in which they've restructured their internal operations. And that is only facilitated when you've got the right technology, isn't it? But I guess fundamentals to all of these things, are often overlooked but so critical to the realities of building these types of great experiences is great data. You know, the old adage of the sort of garbage in, garbage out is still as true as ever before, right? But how do you get access traversing those systems, traversing those channels, you know, those multiple back-end systems that people have? getting access to that consistent data to have, I think you said earlier on, like a digital twin for a customer.

32:53David Brear:That's so key to this, isn't it?

32:56Jouk Pleiter:Yeah. So historically, Backbase already, like the banking OS has an integration layer and we basically industrialize like more than 80 different connectors. So out of the box, you can connect to all the major core platforms. You know, there's like dozens and dozens of fintechs like with payment gateways, with KFC, with AML. So it's almost like an industrialized number of plugs to make sure that at least the basic plumbing to bidirectionally, you know, connect your engagement layer or your banking OS with any underlying legacy banking system or the major ones first and the fintechs. I think that integration layer by default is in place in order already to drive, for instance, mobile and online banking.

33:36Jouk Pleiter:But to your point, it normally was always used as a, you know, we call it pass-through architecture because you didn't want to really persist all of that data in the channel layer. It has to stay in the system of record or in the data lake. So I think we're now with the AI revolution, we are evolving from that and we say we actually are going to persist, but we're not going to persist. So we are not, the banking OS is not the system of record. You will have many systems of record, but Backbase is not a system of record. It's also not your data lake. But what it is, we want to have the truth layer.

34:06Jouk Pleiter:And the truth layer is really important because customers need to understand the truth about them. So what is the truth? The truth is basically, can we come up with something where we can get data sets from 50 different sources, from the credit bureau, from the KYC, like all of those things. And how do we kind of inject that into, we call it an ontology. So ontology is basically business language for like, let's describe the customer. Like, so there's a person, this person can be a customer in retail. It can be an employee. You basically start to model all the different actors in the front office.

34:38Jouk Pleiter:And everything in the front line has now basically a similar ontology. Very much inspired by Palantir, what they're doing in their data aggregation world. We're now translating that into banking and basically operational data in the bank to run the bank and to run the workflows. So you've got ontology. You basically define your language. Then you have a customer state graph where you know exactly for a customer any type of thing you want to know about that customer in that moment. And you can also, and this is very important, you have a context graph where you can basically store any historical decision you're making around that customer.

35:12Jouk Pleiter:Let me give you a very concrete example. You're onboarding a customer. You're doing a KYC check and there's an outcome. Normally, you don't store that decision or like the data points for that decision. You know, it's green. If it's green, you basically onboard the customer, but it's gone. Ideally, you want to store that decision. Then you do the AML check. You want to store that decision. So these are just little examples. But throughout the journey and all these critical workflows, you're making a lot of decisions. That decision is never getting stored in the core or in the CRM, just the outcome of the decision.

35:43Jouk Pleiter:So what you want to do is basically you want to aggregate, A, in one ontology, one customer state graph. So anyone, like the relationship manager, the customer, thousands of agents, tell me the latest about this customer. And not just the reality of core one, but basically aggregated across the bank. That's the problem. So I think your point about data unification, I think is that we don't need to add another data lake. We don't need to add another system of record, but we need to invest in a truth layer that can be used for agents and people to reason on. I know it's a pretty big ambition, but I basically imagine a future where you have 200 agents or 2 ,000 agents and you give them a subset of the data.

36:27Jouk Pleiter:Well, I hope they have enough data to make the right decision for you. or the right historical context. So if you really think about agents and scaling it to full autonomy and to the maximum efficiency point where you can drive it, you need to have a truth layer. Yeah. And I think it's already proven with Palantir and others in other industries, but I think we need to have something like the Palantir of banking. So that's one big component of the banking OS, just to make sure that it's not about more data, but bringing it together to something you can reason on.

36:59David Brear:100%, yeah. I really like that. It stops it being a purely, I mean, don't get me wrong, the benefits from an operational perspective and from an experiential perspective are huge. But actually, you know, anybody out there listening to it who sits in a regulator or in a, you know, a regulatory or a compliance function, all of the things that you've just said are absolutely music to their ears in terms of being able to, it's like having a black box for every individual customer because you can play back the experience, you know, the data, the context, what led to that decision. that's a different level of understanding.

37:34David Brear:You know, KYC used to be something we did before we gave somebody a product. But if you truly want to get to know them as an individual, then you need exactly this type of thing. I guess that really, you know, takes me to what we were saying earlier on. This is a sort of an evolution in the technology set. And I know, don't get me wrong, AI is a massive leap, but it's doing things we've done and moving them and make them happen. But it's a bit of a revolution when it comes to the cultural mindset within those organizations. So which do you think is going to be harder for big banks to adopt? Is it, you know, the technology I think they can evolve and understand, but those leaps in cultural change seem pretty huge.

38:17Jouk Pleiter:Yeah, the way I think about it, like at least maybe two elements to the answers, like two different perspectives. The first thing is kind of, hey, I really think we need to have a technology stack update where at least in the context of Backbase, we already had the integration layer with the digital banking layer, with the workflow layer, et cetera. But like you need to add an agentic runtime, critical infrastructure, like and how to make that compliant. You need to have a, so you need to have an agentic layer. You need to have a truth layer, the ones we just described. You need to have a whole new idea about policing and guarding thousands of agents on permissions.

38:49Jouk Pleiter:We're going to discuss about that, like that's critical. And you need to have a layer where you can basically train change small language models, build your own algorithms, use algorithms. But like, so I think in the technology stack, you really have to see like, we had a pretty relatively straightforward stack with mobile banking. I think now that stack is getting way more sophisticated. And I think that's also what platforms like banking OS are fulfilling, like a wide label set, like, let's make sure that from the get-go, your digital banking platform is also ready to scale very aggressively into an AI future.

39:23Jouk Pleiter:And the second answer on the culture side, I think the big blocking factor is basically from a regular point of view, regulatory point of view, like people don't know how to kind of get the sign off. I really think that is the big unlock that is needed because that's where you stay in pilot mode and that's where you're experimenting like, okay, we're making meeting summaries. Okay, congratulations, right? Like how do we really drive it into our core, core, core process? So I think what we call governed execution, let's prepare for an architecture where we can have thousands of agents. Agents that are provided by Microsoft, by your CRM provider, by your core provider, by Backbase.

40:04Jouk Pleiter:Like, you know, it's going to be a multi-agent world from multiple sources, right? Partners, internal, external, what you have, right? How do we start to kind of put the governance and the compliancy around that? that's kind of like, that's why we introduced Sentinel. Like, I think it's going to be a critical piece of technology. Like, how do we make sure, like, know your customer, know your employee, now know your agent. So where do you register agents? Where do you define their permissions? Where do you define the policies per subdomain in banking, like in underwriting, like the SOPs, the, you know, the pre-checks, the post-checks, like all of those things.

40:39Jouk Pleiter:And ultimately, how do you build the agents that fully observability about what's going on? Everything, every decision. Like a thousand agents calling 500 million API calls, all 500 million API calls, how do I register it? How do I approve it? Like this is the difference between OpenClaw and nobody knows what's going on versus the power of OpenClaw for maybe people in the audience is basically sandboxes for full autonomous agents just figuring it out together. A nightmare in banking, of course, in a regulated industry, complete nightmare. You wouldn't even consider it privately at home, right?

41:15Jouk Pleiter:So, but like that, I think we can solve that. And I think we need to solve that. So how do we build in the guard function like that you can have thousands of agents, but they are controlled by deterministic logic that is guiding them and giving them the context in which they safely can operate. I think if you don't solve that, I don't see AI at scale in banks at all. And I basically see the risk and compliance arguments paralyzing innovation. So I think that's kind of probably the game changer here. like you need to be here, I think, in five years. So how do we kind of get the risk and compliance team in the team, in the project execution on day one?

41:54Jouk Pleiter:And they're basically helping the finding and setting the guardrails and the policies and everything. And then they're helping the agents that will help them to automate the reporting to the regulator. So I think that's the, it might like, everything is stuck in pilot mode, at least for the serious stuff.

42:10David Brear:Definitely. It's the brave people who push beyond that and actually start putting things in the hands of customers, isn't it? It's a level of experimentation beyond really what we've seen. I think if anything, if you look back through the years with financial services innovation, then the speed of it is changing so dramatically. Those cycles between, as you say, checks and plastic to internet banking and mobile banking, the dev kit that we've got each time is dramatically improved. But it feels like we're moving into, you know, light speed now in terms of what it can deliver. Ultimately, I think the promise of the internet was, you know, a deeply personalized, deeply engaging one-to-one experience using data, right?

42:58David Brear:And it feels like finally we're at a point where the technology has probably caught up with the potential of kind of where we get to. But what do you think if this continues, you know, if this direction of travel continues for the next five years, 10 years, what does AI native banking actually look like in that future? Because it has such knock on effects for, you know, customer experiences or the time we have to spend doing banking or fundamentally banking business models in their own right. I think there's probably a whole other hour of conversation we could do on that one if you want to come back and talk about it.

43:34David Brear:but it feels like quite an exciting new world, doesn't it?

43:38Jouk Pleiter:I mean, yeah, no, totally. And the amount of change, like I will take a very, let's say, ambitious stance. And I think it's inspired by some of the work Elon is doing and like where he wants to go. But like, I think we can generate a bank that can service 50 million customers with 50 people. Wow. And of course, I'm now really basically, you know, annoying a lot of people here in the call and like it's impossible. But think about the most extreme format. I think the power, the issues, look at the speed what you see with Entropic. I think it's safe today to conclude that the models are not the bottleneck.

44:16Jouk Pleiter:So that's absolutely not the bottleneck. I think it's now at the application level, at the application layer, how do we kind of make sure that we can tap in these super powerful models and how they can help us with software development, with software implementation, stuff like banking OS. I think in the first time, we now have this raw power where normally you had to hire 500 people from Accenture with a so-so outcome or like, no offense to a Mexican, like in general, any system integrator, right? Very difficult. Everything was like almost like career limiting move because modernization is so difficult.

44:49Jouk Pleiter:I really think we're going to go into an era where also with agentic software development, the modeling platforms like BankingOS, I think we're going to go into a world where you model this. Agents will help you to kind of figure it out within the boundaries of the platform. So I think we're going to see superpowers to actually modernize banks at scale into this new operating model. So the funny part is we're not only developing banking OS, but we're developing, we call it factory OS, which is basically another agentic platform on top of it just to help them build the machine. Does it make sense?

45:18It's almost like the machine building the machine.

45:21Jouk Pleiter:We're entering in that era. And therefore, I think we have to get used to an operating model where things have to go 5x faster. I cannot say banks will get there, right? But I think your worst enemy, and I don't think it's in five years, but someone is going to run a 50 million customer. Someone is going to do Revolut, but then time tens, and then run it with a fraction of the cost. Because if you truly, truly can do governed execution, agents can do most of the work. Really, they can do most of the work.

45:54David Brear:Yeah, which is a very exciting potential for transforming how customer experiences are delivered. But I guess in that future of banking in an AI native and built around that unified frontline, like you say, what's the one thing you'd advise banks to do today to avoid being left behind? Because there's so much noise, there's so much trepidation of acting or waiting or tiny little pilot here or tiny little pilot there. But what would you urge people? the CEOs of banks listening to this now, what do you urge them to do?

46:30Jouk Pleiter:Yeah, I think maybe mentally just declare like, this is the most aggressive change management you probably ever will do in your life because you're basically fighting gravity, which is kind of, you know, the status quo of the past. And like, first question is, are you a change maker? Second question is that, are you still have energy to change? But like, if you, then I would go very, and I meet a lot of people that are 100 % convinced we have to go now. And I think, yeah, the playbooks are emerging. I really think the playbooks are emerging. And I think it's not with hyperscalers and it's not about tokens.

47:02Jouk Pleiter:And I think that's all horizontal infrastructure. It's not solving, it's important, don't get me wrong. But you now really need to start to think about, okay, operating model change. This is like where mobile was maybe still, it was partly operating model, like lower cost to service and let's move another channel. But I think in a way, everything other stayed in place. So it didn't materially change a lot of other parts of the bank, was just adding something to it. I think this one is really a fundamental operating model change where the thing is now connecting the dots. That I think kind of that overlay of a control plane that brings stuff together.

47:37Jouk Pleiter:And if you are in that movie and you believe that is true, I think then you can make it super pragmatic and small and you can basically fix it one journey at a time. I would just, we have consultants, they basically do, we call it an Ignite workshop. In my mind, what they do in a couple of days, they just do an MRI scan of all your journeys. They just do intent fulfillment. They ask you for all the key metrics, like waiting time, et cetera. You can do it for consumer banking and retail banking, wealth management, et cetera. And within two weeks, you've got the full MRI of where you have value leakage, right?

48:08Jouk Pleiter:And based on that, you can basically pick your pilot project and you can actually... And of course, you need to get sign-off on architecture. You need to get sign-off on compliance. So I think it's more about having the intent to change than doing the MRI scan to basically bring the roadmap into very tangible ROI-driven chunks. Yeah, and then get going.

48:29David Brear:Nice, very good. I always say on that, as the beginning of any good 10-step process, probably the first step is admitting that there's a problem, right? And the world's changing so dramatically in this world that with all of the advancements on technology and new competitors, then I think it's pretty easy for people to take that first step sometimes, just admit there needs to be change and start embracing that in that sense.

48:55Jouk Pleiter:There's a lot of people that will change, David. At least maybe I'm in a bubble, but there's plenty of people that are very brave and very active.

49:01David Brear:Definitely, definitely. I completely agree. And it always takes the people who are early to, you know, somebody who was, the followers need a leader, don't they? But we could talk to you for hours on this and we'd love to get you back on as this develops and as you've got more and more momentum around this type of change. but we're going to have to wrap up the show, I'm afraid. So thank you so much for joining me. Where can people learn a little bit more about all the great work you're doing at Backbase and all of the content that you're putting out now as well?

49:32Jouk Pleiter:Well, you can find my name, of course, on LinkedIn. That's maybe very easy. Or backbase.com. That's it.

49:39David Brear:Very good. As for me, drop me an email, david at 11fs.com. Thank you so much for listening, everybody. If you like what you heard, follow the podcast. Don't forget to leave us a review. It helps us make the show better and helps other people find the show as well. As always, if you want to join the conversation, you can find us on pretty much every social media channel at this stage. Just search for 11FS or FinTech Insider. Or if you really want to, email us on podcasts at 11FS.com. Really appreciate you listening, guys. Speak to you soon. Goodbye.

50:11Jouk Pleiter:Some follow the noise. Bloomberg follows the money. Whether it's the funds fueling AI or crypto's trillion-dollar swings, There's a money side to every story.

50:22David Brear:Get the money side of the story. Subscribe now at Bloomberg.com.

50:30Jouk Pleiter:There's a new way to sweet green. Meet Wraps. Handheld, hearty, and made for life on the move. With bold, chef-crafted flavors, fresh ingredients, and over 40 grams of protein, they're built to satisfy without slowing you down. Try wraps today in the app or at order.sweetgreen.com. Available at all participating locations.