# What 3.2 Billion AI Conversations Taught Bank of America

- **Show:** Banking Transformed with Jim Marous
- **Published:** 14 May 2026
- **Source:** https://thefinancialbrand.com/banking-podcasts
- **Transcript:** https://podcasts.opennash.com/shows/banking-transformed/what-3-2-billion-ai-conversations-taught-bank-of-america-57e9e2/
- **Provider:** OpenNash Podcasts (machine-generated)

## In brief

Bank of America's virtual assistant Erica has grown from a simple question-answering feature into shared infrastructure for the whole bank. Jorge Camargo argues that 3.2 billion conversations showed what customers actually want, and that the real value lies in timing: being proactive and helpful at the right moment, not just answering questions.

The central tension is between moving fast and staying trustworthy. Camargo describes a 'battleship, not a speedboat' approach, in which legal, risk and compliance teams design alongside engineers. That lets the bank make tens of thousands of changes while moving from answers, to actions, to eventual autonomy, and from a destination chatbot to AI embedded everywhere.

## Key insights

1. **Customer conversations collapse into about 700 topics** Camargo expected complexity from 3.2 billion conversations but found clarity: consumer banking boils down to roughly 700 topics. That let the bank assign subject-matter experts and write better answers.
2. **Most Erica conversations are now proactive, not reactive** Only about 35-40% of conversations are the classic ask-and-answer loop; around 60% and growing are Erica offering something first. The underlying customer question is 'Am I good?'
3. **Compliance teams co-design instead of approving at the end** Legal, risk and compliance partners joined from day one, which made 85,000 changes since launch workable. Waiting until the end would have buried the team in paperwork.
4. **Requiring Erica integration multiplies the platform's reach** Every new digital feature, such as a credit card or mortgage tool, must include a go-to-market item for how it connects to Erica. This spreads Erica's capability across every team building software.
5. **Smaller banks should copy the discipline, not Erica** Camargo says technology is now widely accessible and may even let smaller banks leapfrog. He advises splitting work into what people do uniquely versus what AI does uniquely, and reinventing processes in an AI-native way.

## Questions this episode answers

### What is Erica at Bank of America and what does it do now?

Erica is Bank of America's virtual assistant, launched in 2018 as a way to get quick answers. Camargo says it has become enterprise infrastructure supporting consumer banking, employees, wealth management advisors and treasury sales.

### What did Bank of America learn from 3.2 billion Erica conversations?

Camargo says customer needs boil down to about 700 topics, and that clients mostly want reassurance that everything is okay. This pushed the bank toward proactive messages, which now make up around 60% of conversations.

### What does 'agentic' AI mean in this conversation?

Camargo describes three stages: answers, then actions the AI performs for you, then autonomy, where AI acts on its own, such as moving money. Actions are arriving now, while he thinks autonomy is five-plus years from being welcomed by clients.

### How does a big bank avoid AI mistakes while moving quickly?

Camargo says the bank is a 'battleship, not a speedboat', aligning legal, risk and compliance from the start and planning for things going wrong. As an example, Erica was quickly updated with existing assistance information during the Los Angeles wildfires.

### What should a smaller bank do first with AI?

Camargo says don't copy Erica; copy the process and discipline behind it. He suggests imagining an all-capable AI, splitting tasks between people and technology, and rethinking business processes. Marous adds that starting with one narrow area is wise.

## Facts

- **Jorge Camargo** [0:00:58]: "We started almost 10 years ago, Jim. Launched eight years ago."
- **Jorge Camargo** [0:01:27]: "I've been building consumer facing software for 25 years. Most recently, eight years at Bank of America. Before that, eight years at the Walt Disney Company."
- **Jorge Camargo** [0:02:40]: "into the solution that is now helping us kind of do the work of 11 ,000 people."
- **Jorge Camargo** [0:03:37]: "we've had about 3.2 billion conversations that have come through Erica."
- **Jorge Camargo** [0:03:37]: "you could really boil down everything in consumer banking at a company like Bank of America to about 700 topics."
- **Jorge Camargo** [0:04:39]: "The way we look at Erica is first, it's all about answers. Then it's about actions. And then finally, it's gonna be autonomy."
- **Jorge Camargo** [0:05:06]: "we think are five plus years out to really be welcomed and adopted by clients."
- **Jorge Camargo** [0:06:43]: "Today, that loop is about 35, 40 % of our conversations, fully 60 % of our conversations and growing"
- **Jorge Camargo** [0:07:45]: "The first generation of Erica that we've built was built with technology from eight plus years ago"
- **Jorge Camargo** [0:08:52]: "In any given month, we'll have between 1.4, 1.5 billion logins to our mobile apps, right?"
- **Jorge Camargo** [0:11:41]: "Perfect example, it's how do we connect all of our core deposit platform systems or our credit card systems."
- **Jorge Camargo** [0:13:31]: "one of the ways that we use Erica is to help forecast, for instance, your balance, right?"
- **Jorge Camargo** [0:16:07]: "in banking, like 95 % of the interactions that you'll have with your bank are very, very transactional, standard, lower stakes, right?"
- **Jorge Camargo** [0:18:12]: "Erica will even help you check in once you go to the financial center and pass that information back to your banker"
- **Jorge Camargo** [0:20:13]: "we have this team of data scientists, Small But Mighty, they have made 85 ,000 changes to Erica since we launched."
- **Jorge Camargo** [0:21:59]: "Once you launch a new digital feature at Bank of America, one of your go to market checklist items is what is your integration back to Erica?"
- **Jorge Camargo** [0:23:37]: "we recently announced kind of the relaunch of our rewards program from preferred rewards to Bank of America rewards."
- **Jorge Camargo** [0:28:40]: "Although 2 % of our Erica interactions are voiced today."
- **Jorge Camargo** [0:29:48]: "every three months we'll remind you, hey Jim, here's the people that you have given authorization to use your data."
- **Jorge Camargo** [0:33:56]: "I had the opportunity to step into the Erica team four years ago."
- **Jorge Camargo** [0:37:27]: "we have a mobile app for Merrill Lynch. We have a different one for our robo-advisor Merrill Edge"
- **Jorge Camargo** [0:37:54]: "we essentially brought five apps together into one and delivered in a single package."
- **Jorge Camargo** [0:38:22]: "we have that distribution that reach 65 million clients"
- **Jorge Camargo** [0:39:50]: "I was in San Diego for the Consumer Banker Association's annual meeting, CBA Life."
- **Jorge Camargo** [0:41:37]: "assume there's an AI so advanced that it can do anything. Then peel that back."

## Transcript

**[0:00:11] Jim Marous:** This last session is with Jorge Camargo, head of digital platforms at Bank of America. And the title is kind of intriguing, but it's how 3.2 billion conversations, what they taught Bank of America. And we're going to go deep into how important those conversations are, and probably give you a little idea of what some of the really progressive organizations are using their conversations for. So welcome, if you can, Jorge Camargo from Bank of America. So there's a lot of different ways we could have gone with this conversation.

**[0:00:44] Jim Marous:** But the best seems to be, you know, personalization. Because if there's one thing that everybody knows about Bank of America, it's that you have a very progressive platform that was built.

**[0:00:58] Jorge Camargo:** how many years ago now, Erica? We started almost 10 years ago, Jim. Launched eight years ago.

**[0:01:03] Jim Marous:** And it started as simply something that you could ask some simple questions for, the things that now at a call center has been outplaced anyway. But it's gotten much more important. And you are in charge of digital platforms. So let's start from the beginning. Your role and your roles before Bank of America, where have you been?

**[0:01:27] Jorge Camargo:** Well, first of all, thank you for having me. Thank you. Long time listener, first time caller. I've been building consumer facing software for 25 years. Most recently, eight years at Bank of America. Before that, eight years at the Walt Disney Company. So I've been at this intersection of the internet, mobile, now AI. And how do we just create these world class client facing experiences?

**[0:01:52] Jim Marous:** So Disney. So I'm starting to think that this may have a trend here with regards to what our conversation is going to be. But when Erica launched in 2018, what's the difference between the vision then and what the vision is now?

**[0:02:10] Jorge Camargo:** So what's fascinating is the vision hasn't really changed. But when we launched Erica, it was more of a feature. It was a solution for a customer, right? How do we get you to the answers you need in the fastest possible way? And when you step back and look at Erica today, it's become like enterprise infrastructure that's powering not just our consumer bank and our consumer bank experiences, but it's spread like wildfire. It's now helping us support employee conversations.

**[0:02:40] Jorge Camargo:** It's helping us support wealth management conversations with advisors. It's even helping us in our business to business relationships with treasury sales. So that's a fascinating evolution of Erica kind of like growing from this initial idea, small scrappy team into the solution that is now helping us kind of do the work of 11 ,000 people.

**[0:03:01] Jim Marous:** So 10 years ago, you couldn't have envisioned where the world would be today. We could have dreamed about it. We may have thought it was going to come earlier or later or even where this is going to go. What does Erica provide us? What do those conversations provide you in a world that we're talking about, at least, agentic AI, having agents work on customers' behalf on an individualized basis, based on their needs? What did all those conversations do to prepare you for what comes next?

**[0:03:37] Jorge Camargo:** Yes, I think there's two parts there. The first one is we've had about 3.2 billion conversations that have come through Erica. And first of all, that's humbling, the amount of just trust and conversations that happen. But where we thought we'd get complexity, we actually got clarity. We really understand what do our clients want to talk to us about. And the fascinating kind of takeaway was you could really boil down everything in consumer banking at a company like Bank of America to about 700 topics.

**[0:04:10] Jorge Camargo:** Only 700? Only 700. But listen, it's more manageable than organize the world's information, right? But once you really nailed that down, you can really then focus on, okay, here's what clients are interested in, here's what we need to talk about. Who are the subject matter experts in our organization that can help us do that? And then you can really build on top of that and really create better answers. And to your second point on agentic, right?

**[0:04:39] Jorge Camargo:** The way we look at Erica is first, it's all about answers. Then it's about actions. And then finally, it's gonna be autonomy. So we've gotten really good over the past eight years at giving you the right answer and really being able to route you where you need to go to get that answer you need. We're getting better at providing those actions. So this is what agentic is all about and why we're so excited about the possibilities there.

**[0:05:06] Jorge Camargo:** All of a sudden, things that weren't economically feasible in the past to automate, all of a sudden agentic technology gives you the ability to say, hey, we can do that for you. You don't have to do it yourself. And then finally, the autonomy piece. I think that's further out when you really look at this notion of like a digital twin or AI moving money on your behalf. Certainly things that we look at, but we think are five plus years out to really be welcomed and adopted by clients.

**[0:05:35] Jim Marous:** So the 3.2 billion conversations is really not just about offloading people. It's not just about volume anymore. It's about what you do with that data. What are some of the things that those conversations have enabled you to do that maybe was part of the plan, but probably scaled at a much bigger value? And what it gives you the foundation for? Without giving away trade secrets, what does this really prepare you to do and what has it already enabled you to do?

**[0:06:05] Jorge Camargo:** Yeah, absolutely. So the most important thing that we've learned along this journey is clients come to us for these answers, right? Like very punctual. Here's my question. Give me an answer. But that doesn't happen that often, right? Most of our clients are very comfortable with their banking relationship, with their digital solutions, with their digital tools. What we really discovered is there's this overarching question of, am I good? Is everything in my world okay? And that goes across the spectrum, whether you are just trying to get out of debt, to whether you're worth millions of dollars.

**[0:06:43] Jorge Camargo:** So what has really taught us from a development perspective, from a building perspective, is how do we start getting, not away from answering your question, but once we answer that question, how do we start trying to preempt or to be proactive about information that might be valuable to you and answer that question of, am I good? So for us, Erica started as a, ask me a question and get an answer. Today, that loop is about 35, 40 % of our conversations, fully 60 % of our conversations and growing is more of this preemptive, proactive, like, hey, let me tell you something that I think you might be interested in that might be valuable to you.

**[0:07:26] Jorge Camargo:** The so what's. Exactly. So in that evolution of the product,

**[0:07:33] Jim Marous:** do you see it transforming into not just here's something you may want to know to asking questions back to get deeper information, deeper insights?

**[0:07:45] Jorge Camargo:** Absolutely. And I think this is where the large language model conversation comes into play. Being able to have much richer, much more interactive conversations. The first generation of Erica that we've built was built with technology from eight plus years ago, which in AI years, it's like dog years, right? So it was very much focused on, let me try and match your question to the best possible intent or answer that I have. Now, with the power of language models, you're able to really engage in a much deeper conversation and bring in a lot of more data points, have a lot more kind of like branches to your conversations, and really mimic more of that interaction you would have in a true

**[0:08:29] Jorge Camargo:** back and forth. So it's an area that we're very, very excited about.

**[0:08:31] Jim Marous:** How do you store all this information? Because it's not like we have this many customers that called about balances. That's not the good data. It's okay, but it's not good. But how do you store it so that, does Erica remember that I visited you with a concern three weeks ago and I'm calling again about the same concern? And how do you store this?

**[0:08:52] Jorge Camargo:** Because it's got to be a lot of memory. Listen, it's a world-class technology team. When you look at Bank of America, the investment and the focus that goes back to resiliency, availability, performance, it's massive, right? We have tens of millions of customers. In any given month, we'll have between 1.4, 1.5 billion logins to our mobile apps, right? Wow. Which is just a mind blowing number. And there's a whole team of just very talented technologists that are focusing on how is that experience always resilient, always available?

**[0:09:29] Jorge Camargo:** Because trust is obviously critical for any financial institution. You want to be able to access your digital solutions. You want to be able to take action. So we always want to make sure that that information is there, it's available, it's reliable.

**[0:09:43] Jim Marous:** It's supposed to be a great tool for personalization, make it easier. You know who's calling you, you're going to be dealing with their issues. However, as an industry, as all industries are challenged by, the integration of channels is the hard part of this. because they're all based on different aspects, based on human versus digital, digital versus human, branches versus call center, all these issues. And I have a case study that some people have heard more than once probably with Delta Airlines where I had a challenge.

**[0:10:17] Jim Marous:** I first went to the human on the plane because that was easiest accessible. They didn't get me where I wanted. I then started because they give you wifi on the plane so you can start talking to a chatbot or something real quickly. That didn't work out the way I wanted to. I found another channel. And as sometimes happened, when you're getting really frustrated, you finally decide, I'm going to do the nuclear warhead. I'm going to go to social media.

**[0:10:39] Jim Marous:** And they quickly say, let's take this offline as quickly as possible. That being said, how are you dealing with being able to meld the insight you get from Erica, from the insight you get from the digital and human aspects of your call center, to the insight you get from branch interactions? how do you how are you dealing with that major challenge because it's because delta by the way at the end of that story they called me and said i think we do pretty good by channel we do a terrible job of integrating that's what we're working on right now it's the holy grail yeah

**[0:11:15] Jorge Camargo:** and i think a lot of large enterprises are struggling or going through this transformation is how do you connect all the different pieces of the puzzle how do you assemble it together The beauty of a system like Erica, the way we've built it is, there's a lot of piping that goes underground in trying to bring as much information, as much context as possible. So that when you engage with Erica, Erica has a little bit of that context of what's happening.

**[0:11:41] Jorge Camargo:** Perfect example, it's how do we connect all of our core deposit platform systems or our credit card systems. Let's happen you have a declined card transaction. So that was a lot of effort to connect that credit card system to feed that information into Erica so that whenever you open that interface, it's going to be able to preemptively ask you. It's like, hey, Jim, I saw you have a decline in your card. Are you going to ask me about this?

**[0:12:06] Jorge Camargo:** I have some additional information for you. Or let me connect you to an agent, right? Because that might be one of the things that you're asking about.

**[0:12:14] Jim Marous:** So personalization is something we all talk about. We implement at a much lower rate than what we say we'd like to. It's the implementation gap that actually happens that we know what we have to do. It's getting there. That's the tough part. What are you doing at Bank of America to enhance? Where's your goal from a personalization aspect to make it so it's really differentiated from all the other organizations that are out there trying to do the same?

**[0:12:45] Jim Marous:** not just in banking, but be it the hotels in Vegas or at home or the car rental company or the airline?

**[0:12:53] Jorge Camargo:** Yeah. So the way we look at personalization, obviously it's a very broadly used word and term. Yeah. We like to think about it as it's not how much knowledge do I have or how much information can I put in front of you, but it's rather a timing question. When can I step in with information that is relevant to what is happening in your interaction at that moment. So what does that mean for us? I mentioned the example of a credit card decline, bringing that into the mix, but it's also about being kind of proactive from a positive perspective.

**[0:13:31] Jorge Camargo:** So one of the ways that we use Erica is to help forecast, for instance, your balance, right? Are you trending in one direction or another based on your historical patterns and subscriptions? Where are you going to be this time next week? So we try to provide that context, that information back to our clients and give them choices. What would you like to do? Would you like to do anything with this information? And that's something that has really kind of built a lot of trust with our clients, being able to really try and anticipate, try to be helpful without being pushy.

**[0:14:05] Jim Marous:** Well, it's interesting because Bank of America, as with Disney, I know we get back to you somewhere. it's interesting because sometimes it doesn't take the master class to make a person feel like the experience is personalized i i referenced to you in the back going i've heard the story i've used it before isn't really true and you said it was that as much as we know that the the bands that we wear for members we sign up for it we'll track where we go and they they know different family structures where you're most likely to go and they know now to the point of prompting you you may be interested in this ride knowing full well it's exactly what you're

**[0:14:43] Jim Marous:** going to be wanting to go to and you may want to sign up for it now but a very interesting aspect was the taking the risk of making it so that um talking characters could actually reference a child by name in an interaction freaking out the parents until the child looks at them and is completely in awe that you know, Snow White knew who they were by name. That's not so hard when you figure out the dots and things, but how do you deal with the too far versus not far enough, the craziness factor?

**[0:15:22] Jim Marous:** Obviously, it's different than it was when Disney first introduced us. We're getting used to this a little bit more, but how are you trying to replicate that level of wow to the customer today?

**[0:15:34] Jorge Camargo:** Yeah, so I'll take off my banker ears and put it on my mouse ears. There we go. And one of the takeaways I had professionally when I was at Disney, it's a company that's completely obsessed with client experience, with delivering an experience that is so memorable that it's almost indistinguishable from magic. And the cast members that work in delivering technology, in delivering experiences, are doing that seamlessly in the background, really kind of toting that line of what is the difference between magical versus might something feel invasive.

**[0:16:07] Jorge Camargo:** So that's part of what I've brought to Bank of America eight years ago. It's like, how do we sprinkle some of that magic pixie dust on some of these banking experiences? And really, it boils down to one of the big, big learnings is in banking, like 95 % of the interactions that you'll have with your bank are very, very transactional, standard, lower stakes, right? Like, I'm going to send you some sell. I'm going to pay a bill.

**[0:16:33] Jorge Camargo:** I'm gonna check my balance. But there's a subset of those interactions that are the high stakes moments that really matter that make or break a relationship, right? And those might be moments of like high stress. Let's say a loved one passed away and you're calling us for support. There might be big commitments. You're trying to send like a wire transfer for a home purchase or for a college tuition check. So it's very important that we connect those signals.

**[0:17:02] Jorge Camargo:** And whether it's a virtual channel or a human channel that these folks know, it's like, hey, you're about to engage in interaction that really matters, that is really going to be important for this person. Know that, have that context, be ready to kind of like put our best foot forward, similar to this Disney example.

**[0:17:25] Jim Marous:** So how are you able to move an anxiety moment on Erica? to a human channel that you say, boy, this is not Erica's bailiwick. And we can't just disengage. We need that agent to say, you know what, I'm not equipped for this. I got to transfer you to somebody else. How does that happen?

**[0:17:46] Jorge Camargo:** Yeah, we have invested a lot in this trust building. So trust is built one step at a time and it can be broken very quickly. So one of the most important skills we've taught Erica is when to step aside and went to kind of go into these off ramps to bring in a human assistance. And from day one, we've tried to understand, hey, is this like, is this contact us? Do you want to reach somebody by phone?

**[0:18:12] Jorge Camargo:** Can we put live chat in front of you? Or is this a situation where maybe it's make an appointment? So Erica will help you create that appointment, pass along the information. Erica will even help you check in once you go to the financial center and pass that information back to your banker and saying, hey, this is Jim, here's what he's here for. So a lot of this focus on creating these off-ramps is because we realize that we want to offer clients that choice, right?

**[0:18:40] Jorge Camargo:** I know high touch, high tech, it's very overplayed. But at the same time, we invest a lot in those high touch off-ramps to make sure that clients know that they have the ability and the flexibility to always reach a human when they need to.

**[0:18:55] Jim Marous:** So that wasn't Erica 1.0. It probably was in 4.0. How do you implement iterations on this platform? Because financial institutions overcomplicate everything and will take forever to get there. Obviously, Erica today and where you see it tomorrow is vastly different from the foundations the same, but what you're able to do and the way you're able to implement it has changed dramatically. The iterations you go through, how does that process work at Bank of America with regard to how they get presented, how they get approved, and how you get it past the whole risk avoidance mentality of banking?

**[0:19:37] Jorge Camargo:** It really is an organizational change and mindset question. The way we lead our teams is by focusing on launching is like showing up to the race. You haven't ran the race, you just showed up. That's when the real competition, the real learning begins once the race actually starts. And with Erica, we've thought about it in two dimensions. One is how do we constantly make Erica smarter? And then the second thing is how do we drive the distribution of Erica?

**[0:20:13] Jorge Camargo:** So for the first example, for making Erica smarter, we have this team of data scientists, Small But Mighty, they have made 85 ,000 changes to Erica since we launched.

**[0:20:27] Jim Marous:** Okay. I don't know about you guys, but when we talk about change management, 85 ,000 changes, I don't know how long that would take a normal organization to do. How long does an individual, most individual changes take? How long from ideation to implementation?

**[0:20:44] Jorge Camargo:** Like these changes are from small to large, right? And the small ones may be typos, right? Like you miss type of words like, oh, this is what you meant. This is where you need to go. All the way to launching large new features. But it requires rethinking a lot of these processes. And there's a lot of partnership from day one. Our legal, our risk, our compliance partners, they were not coming in at the end telling us, no, you can't do this.

**[0:21:11] Jorge Camargo:** Like they were there co-designing with us. And that's one of the big kind of like enterprise aha moments.

**[0:21:16] Jim Marous:** Again, a lesson that we all, most of them have learned it, that you can't leave compliance on to be the last approval because it will never get through.

**[0:21:24] Jorge Camargo:** Never get through.

**[0:21:24] Jim Marous:** But to get them on the front end where they're part of the solution, you can work with them in that way.

**[0:21:28] Jorge Camargo:** And you build a system that scales. Back to your point, it's like just the paperwork on 85 ,000 changes would have just killed anyone, right? But we build from the ground up and saying, hey, we're going to need to every week or so make adjustments, tweak our model, tune it, make, like introduce new answers. How do we do that in a way that's compliant, that's really kind of in line with our responsible growth approach, but at the same time kind of allows us to move how we need to move, right?

**[0:21:59] Jorge Camargo:** So that's the intelligence part. The other part that I always like to talk about is this notion of distribution. So when you launch, again, it's not you're off by yourself, you're trying to get attention. how do you mobilize the entire organization to help you succeed? And that's one of the other secret sauce things that we did with Erica. Once you launch a new digital feature at Bank of America, one of your go to market checklist items is what is your integration back to Erica?

**[0:22:29] Jorge Camargo:** If I'm launching a credit card feature, if I'm launching a mortgage feature, you're being asked, how is this connecting back to Erica? How is Erica providing insights? How is Erica providing answers? And again, I don't want to understate how important that is because now you've multiplied your ability to innovate and grow that core Erica capability to every single other team that's building software in your company.

**[0:22:55] Jim Marous:** So Erica initially was an efficiency play. We can call it whatever it was. It was offloading conversations. It now appears to be transforming itself into much more of an effectiveness and engagement play. What is, give me an example in the last year of some major change that's been made in Erica or capability that's been made that really change is a good example of how it's not just an efficiency play where it's actually using technology, data, insights and deployment.

**[0:23:33] Jim Marous:** to possibly change the consumer's life?

**[0:23:37] Jorge Camargo:** Yeah. So one of our big enterprise initiatives, we recently announced kind of the relaunch of our rewards program from preferred rewards to Bank of America rewards. And that's been a multi-year collaboration and effort. But where Erica comes into play is we have built this whole set of capabilities to identify historically, hey, do you qualify for this program? Like, what are the benefits of this program that you might not be taking advantage of? At the end of the year, here's the benefits that you took advantage of.

**[0:24:11] Jorge Camargo:** Like, here's the kind of the proofs in the pudding of like, here's the fees that were waived and here's the other like deals that you took advantage of. So that's like a fantastic example of like Erica being able to bring forth all these other capabilities that the enterprise is launching. And it wouldn't have been possible if we hadn't kind of built this harness that allows us to tap into all these big initiatives across the company and help kind of shine a light on them.

**[0:24:39] Jim Marous:** You know, it's interesting. Not all of them are as earth-shattering as that and as important as that. But it's not just in the ability to offer that service to customers, but to reference it in conversations. because, you know, I've talked about for years, the thing that consumers want, they want you to know you, understand you and reward you. You cover all three bases right there, but I think it's not always the big deal, but sometimes it's just referencing the fact that they know you that way.

**[0:25:08] Jim Marous:** Yeah, I just had an experience on Delta. I just had a birthday and the seat back message when I logged into the TV said, by the way, happy birthday to him. And then I took the flight back and they said, hope we didn't miss your birthday. And that's little, but I've heard it from, it must be new because I've heard from three people in the last month that they said, oh, Delta, you know what they did?

**[0:25:33] Jim Marous:** And it's the smallest thing. You know, we do it on ATMs and all that, but you see it in the airlines, it's not what you expect. But again, going that step further. Now, there's been some bigger issues from the standpoint of using Erica as the communication and engagement tool for financial wellness. Can you explain some of the things that Erica is now doing to improve not just your, what you might balance, but help consumers along that journey of financial wellness?

**[0:26:03] Jorge Camargo:** And it's a fascinating story because top line, Erica's role, first of all, is to remove a lot of noise so you can focus on the signal. Our experience design team is fantastic And they do all sorts of amazing research with our clients, with prospects, with non-clients. And one of the big insights that they have come back with us is, listen, at every stage of the wealth spectrum, people really want to do three things. They want to monitor their financial picture.

**[0:26:35] Jorge Camargo:** They want to manage or like transact with us. And then finally, they want to optimize, which is like the longer range planning. But the fascinating thing is it's such a kind of complex ecosystem environment that the pyramid, a lot of the work goes to just monitoring, like what's happening or managing, like let me complete all these actions. And very little time is left at the top for you to think about the longer range planning. And what we're trying to do with this whole generation of AI tools and digital tools is to invert that pyramid.

**[0:27:08] Jorge Camargo:** Is we wanna make it so easy via alerts that you don't have to constantly check in for notifications. We want to be able to have self service capabilities and in the future kind of agentic capabilities. So you don't have to spend a lot of time filling out forms or like wire requests, right? Because then that will open up all this kind of ability to focus on these longer range planning items. We have Life Plan, which is a fantastic place where people go and tell us about, hey, here's my objectives.

**[0:27:40] Jorge Camargo:** How do I get there? We have My Financial Picture, which allows you to track your kind of like full set of finances. So really with Erica, what we're trying to do is take out this no joy busy work so that we can really focus or give you the ability to focus on these like medium and longer range planning tasks.

**[0:27:59] Jim Marous:** So is Erica actually making it easier for your customers to work with LifePlan, to do these other things? Because you're able to, unlike in the branch or anyplace else, build that engagement to continually reference where they are in their plan and everything else. Are you, in effect, testingogenic AI solutions in a more voice capability, right? I mean, you're dipping your whole foot maybe up to the waist in water. But is this being used as a planning ground for what we see as the future so that that stepping off point is less dramatic than it would be in a traditional financial institution where they don't have as much data going down that way?

**[0:28:40] Jorge Camargo:** Absolutely. And in some places, dipping our toe, in some places, knee deep. Yeah. Voice, let me come back to the financial wellness. But voice, absolutely, we think that's going to be a very, very important area of focus. Although 2 % of our Erica interactions are voiced today. Interesting. We really believe that that number is just going to continue to grow. But back to the financial wellness piece, The excellent example that I heard recently is we, our clients, are not just focused on the long range planning of like, hey, how will my assets grow?

**[0:29:17] Jorge Camargo:** They're also worried about fraud or privacy or security. So that's part of the comprehensive financial wellness education. And one of the things that they kept asking Erica is like, hey, how is my data being shared? And obviously, at Bank of America, we don't sell or monetize your data, right? Right. But by doing research and going deeper into conversations, we realized that a lot of our clients authorized third parties to access their Bank of America data, right?

**[0:29:48] Jorge Camargo:** It might be a payment app. It might be like a financial tracking app. And they're very concerned about like, hey, what is happening with my data? Obviously, some institutions are very reputable, some of them more early in the process, so they may not have the controls that you need. The point of the story is one of the key insights that we built for Erica is like, every three months we'll remind you, hey Jim, here's the people that you have given authorization to use your data.

**[0:30:19] Jorge Camargo:** You want to go check that? You want to go revoke some? Is everything good? So we're trying to anticipate, again, all those conversations and trying to approach the financial wellness through a variety of different lenses. And again, all in this goal of like, how can we be helpful? Right. And are you OK?

**[0:30:39] Jim Marous:** So that innovation culture runs counter to traditional stability, trust, security, privacy, all these other things. how has Bank of America gotten over that hurdle of accepting relatively different kinds of risk for the deployment of stronger solutions? Because that, we talk about innovation and many times it's a thing over there and, you know, that's our innovation lab. We talk about a lot of this stuff, but you're actually putting on the road a lot of things, more than I even knew based on our conversation today that really would be a hard thing to sell within our financial institutions.

**[0:31:25] Jim Marous:** Obviously, as top leadership allows it and likes that going forward, but how do you hurdle those things that are truly deploying things that most institutions say, we don't want a mistake when mistakes are going to happen? Either someone's going to say something the wrong way, whatever it may be, and you're going to take them down maybe not the perfect path that you would have done if you heard it as a human. How is Bank of America addressing risk rewards?

**[0:31:54] Jorge Camargo:** Yeah. So we like to say internally we're a battleship, not a speedboat. So we don't necessarily focus on being the fastest to market. We always focus on alignment. And as I mentioned earlier, alignment with our legal risk appliance partners from day one is critical because we scenario plan. Like something is going to go wrong. What if this happens? How are we going to react? So back when the L.A. wildfires were kind of devastating Southern California, One of the rapid response abilities that we had was we had trained Erica to talk about our client assistance program and the resources that are available whenever disaster hits.

**[0:32:43] Jorge Camargo:** So for us, that was essentially a very quick pivot to say, now whenever you hear a reference to LA or wildfires, and now let's preemptively go out to some of these clients in these zip codes that we know are impacted. and we're still putting out information that has been reviewed and programs that have been available for a long time. We're just applying that new context to that existing playbook. So that's kind of the planning that we've done historically to address that.

**[0:33:12] Jim Marous:** So you haven't been there for the entire duration of Erica, but I'm sure there's a couple on the team that have almost been there from the beginning. did anybody see erica were they envisioning a platform that would be almost the the front of the boat with regards to how you're going to communicate how you're going to engage how you're going to innovate and how you're going to deploy new solutions out there in the marketplace was ever envisioned as this or is it just sometimes the thank goodness that we came up with this because it really played into our role?

**[0:33:46] Jim Marous:** Because it sounds like it's very much at least central to the organization, even if it's somewhat behind the scenes in some ways. Or is it more front and center more than I know?

**[0:33:56] Jorge Camargo:** Yeah, so I had the opportunity to step into the Erica team four years ago. Some Erica's stepdad, in a sense, have seen her grow into a very prodigal teenager. But some of my teammates who have been there from the beginning, they'll kind of be split. They'll say the vision was always this assistant that's a guide by your side and is getting you the answers you need. But I think nobody in their wildest dreams imagined again, Erica turning into really infrastructure that's running the entire bank.

**[0:34:29] Jorge Camargo:** And that has been kind of very, very rewarding to see how something that they incubated and built from day one is being used in so many places. There's an interesting aspect there as well as, hey, it's not just one team that's building AI and building capabilities. It's how do you build so that all these other teams can take advantage of that? How do you upskill different people, different organizations, so they can learn what you've learned through millions of dollars and thousands of changes?

**[0:35:01] Jorge Camargo:** So there's also a cultural change there for those teammates to become like train the trainers. How do they become like evangelists of what's possible and what can be built? And that's been very, very rewarding for us as well.

**[0:35:16] Jim Marous:** So obviously it's a major innovation engine, but the speed of change is not going slower. It's speeding up. And as far as you're ahead of the game in this aspect with Erica, how do you prepare for what is going to be even a faster evolution to make sure that you stay ahead of the game as opposed to playing catch-up? Because I think all of us would say, well, you have a lot of wind in your sails right now, but how do you keep that wind going?

**[0:35:48] Jim Marous:** Because I had a gentleman here last year from Allied that I said, what keeps you up at night? He goes, what keeps me up at night is the knowledge that eight, maybe 11 people in a garage are going to be able to build what I have today in a three-month period. So that has to be in my mindset. So how do I keep on increasing the speed so I can stay ahead of the payment, the person who's going to make me pay for it?

**[0:36:14] Jim Marous:** How do you do that at Bank of America?

**[0:36:16] Jorge Camargo:** Yeah, the technology world is just moving so quickly. So I live in the San Francisco Bay Area, Silicon Valley's backyard. And just the pace of innovation just continues to accelerate. The last three years since the release of ChatGPT is just, it's a rocket ship, right? Tons of money coming into the space. We meet weekly or I meet weekly with a new potential partner or vendor that is showing you some sort of technology that just completely blows your mind.

**[0:36:45] Jorge Camargo:** So let's just kind of acknowledge that.

**[0:36:48] Jim Marous:** What's interesting is, Eric was built pretty much internally. You almost have to look for the people who are hard paid to be in front of the game, don't you? You have to pick your fastest partners.

**[0:36:57] Jorge Camargo:** Yeah, you continue to evolve, right? But this is where kind of that alignment back to what is our mission, what are we trying to deliver becomes so key. It becomes less about the technology, but like how are you deploying it? And how are you deploying it in a way that's really going to matter to your clients? Like that alignment, I think, eats everything else for lunch, right? Like for us, alignment means how do you remove silos, right?

**[0:37:27] Jorge Camargo:** A couple of examples. When I first joined Bank of America, my remit was to focus on mobile apps and mobile app experience. And the first thing that I faced was we have a mobile app for Merrill Lynch. We have a different one for our robo-advisor Merrill Edge, a different one for retirement plans, a different one for the consumer bank, a different one for wealth management. Yet our most important clients are clients that have multiple relationships across all those brands.

**[0:37:54] Jorge Camargo:** So first of all, our business is, guys, we're shipping our org chart, right? We're not looking at this from a client lens perspective. Clients want one solution, right? They're coming to us because they appreciate having this range of services all in one package. So we created this notion of like a one app. So behind the scenes, we essentially brought five apps together into one and delivered in a single package. Back to this question, how do you stay ahead of innovation?

**[0:38:22] Jorge Camargo:** we're doing something similar with AI, right? We are now embedding AI. It's not just Erica and you go to Erica as a destination, but rather how is Erica now going into every little nook and cranny of our digital experience? How is Erica going to go into every nook and cranny of our associate experience? And then how do we have these AI capabilities that are creating AI-enabled clients, AI-enabled associates? and we believe that that's where we're going to really be able to to excel right because we have that distribution that reach 65 million clients right you put the right tools in front of those clients right and and that's when you really get kind of like take off it's

**[0:39:06] Jorge Camargo:** gonna be pretty fun so

**[0:39:09] Jim Marous:** organizations here i'm gonna assume i'm gonna take a wild guess that there's no institution that's larger than Bank of America. In fact, you probably have in any one state probably more assets than most institutions here do. If you were to move to a smaller institution from where you are today and be assigned to make this happen, what's the first thing you'd recommend these people do or what's the first thing that you would do if you went to organizations that does not have Erica?

**[0:39:42] Jim Marous:** Where should we start on the personalization and the AI journey in order to at least try to tread water until we can catch up?

**[0:39:50] Jorge Camargo:** Yep. So I wouldn't copy Erica. I would copy kind of like the discipline and the process that got us there. So two weeks ago, I was in San Diego for the Consumer Banker Association's annual meeting, CBA Life. I was a digital member committee there for five years. And what was fascinating is it's not the large banks that are the main beneficiaries there. It's all midsize, the regionals, the smaller local banks. So what was very enlightening about those experiences at CBA, not just this year, but all the years prior, was to really understand the point of view of what a smaller institution is facing.

**[0:40:31] Jorge Camargo:** Now, back to the technologies moving fast. Any institution today is going to be able to have access to the technology, potentially even leapfrog the technology that we've just spent over past eight years, right? So it's not about the technology. Again, it's how do you put that in front of your clients? How are you going to deploy that internally? How are you going to support that internally and make it work for you? So again, the tech disruptors are here, right?

**[0:40:59] Jorge Camargo:** They're going to put these tools in the hands of a lot of different people. It's all about what you do with those tools.

**[0:41:05] Jim Marous:** So I'm going to double down on the question and go, what is the one thing they need to do to be prepared and to move forward in a world of Erica's eugenic AI, all the other things. I mean, obviously I have to get my data shop in order. I have to also look at the data shop from the point of deployment because insights alone mean nothing if you don't find a way to deploy them at scale.

**[0:41:28] Jim Marous:** How do I ready an organization and make it so that we can start making impacts? Because we can't even wait till the

**[0:41:37] Jorge Camargo:** end of the year? What do we do? So the thought exercise that I conduct internally is assume there's an AI so advanced that it can do anything. Then peel that back. What are the things that are going to be uniquely done by your associates versus what are the things that will be uniquely done by technology? Like literally create two buckets. Yep. And then go operate based of that, go change the ability that a smaller company or a smaller bank would have is can you be a little bit more radical in re-envisioning your business processes because i think that's what ai is all about is is reinventing your business processes it's just one more technology piece so if you can

**[0:42:19] Jorge Camargo:** bring ai and really reinvent how that business process is run in an ai native way that's where i think you can really create some some long lasting advantage

**[0:42:29] Jim Marous:** I agree. I think to narrow your scope of what you're going to solve for first, because it can be overwhelming as we know in generative AI or any interaction you have with the generative tools, narrow the focus and get that right. So you can get your case study to deploy more, because it's going to take more investment. But if you can say, I'm going to make it so this part of my organization gets it right. And I'm going to, it may be credit card and maybe in the new account checking account area maybe in cds maybe in investment services and narrow it so you can make an impact and use the low hanging fruit as your your starting point jorge

**[0:43:07] Jim Marous:** i can't tell you how much i appreciate this conversation i've honestly been wanting to have a conversation with your team on this for probably the last five years because i saw that you what's needed is you know what's coming up your train's moving and it's it's kind of interesting to see what kind of head start you have. But as you said from the very beginning, it's not that you can't jump really quickly because we don't have to go to the first steps of simply looking at saying, what's my balance?

**[0:43:34] Jim Marous:** This is my balance. It's a lot bigger than that. So thank you very much. Thank you, Jim. Thanks for listening to Banking Transformed, the winner of three international awards for podcast actions. If you enjoy what we're doing, we would really enjoy a positive review. Also, check out my recent articles in the financial brand, the research we're doing for the Digital Banking Report. This has been a production of Evergreen Podcast. A special thank you to our senior producer, Leah Haslidge, audio engineer, Chris Fofaglias, and video producer, Will Pritz.
