Introduction

Doug English: [00:00:00] Steve O’Donnell and Saroop Bharwani, welcome to CU On The Show I’m delighted to, to have you with me today to help us capture the boldest ideas with AI in the credit union movement.
So, uh, I, I’d love to start, uh, uh, hearing from either of you, maybe, Saroop, you’re in the top of my screen. So you go first, tell us how did you get started working with credit unions and maybe a little about those pictures behind you.

Saroop’s Background & Inspiration

Saroop Bharwani: Yeah, absolutely. Thanks for having me, Doug. Uh, pleasure to be here. Uh, so I Am the, co-founder of, of Senseo, um, most of my career has been working with, uh, my founding team on building like AI products for the financial services industry.
I’m from Canada, so like the natural thing to do in, in Toronto is working for the big five banks, and that’s kind of where I, I got my, you know, school of hard knocks, uh, learning about the financial services industry. I was building large scale ML models for years, and in about 2021, uh, I met the great folks at CU two who introduced me to great guys like Steve, and, uh, really fell [00:01:00] into the credit union movement and loved the centricity of just being focused on people and serving for people, and really creating impact in that way.
And, um, post chat GBT, the world sort of exploded. Um, you know, everybody wanted to know about AI and had the pleasure of meeting, uh, Steve and a number of other credit unions where we built a very industry specific product. And, uh, little bit about these pictures. Uh, when chat GPT launched, I, uh, started a, uh, a podcast call it, or a webinar, um, in early 2023 called Fine Tune Fridays.
And I actually used Midjourney to create these pictures, uh, on, on the call alongside credit union. So these are who I, uh, uh, who I feel are like heavy contributors to where we are today when it comes to ai. Uh, Johannes Gutenberg, the founder of the printing press. You know, language is a very big thing there.
We got Nicola Tesla, who had a very big part in electricity. Electricity is the foundation of, of enabling compute and large scale, uh, pre-training. Uh, we got, um, [00:02:00] uh, we got Turing over there, the inventor of the computer, and we got Steve Jobs who, you know, the personal computer. We wouldn’t be on this podcast without that.
So that’s a little bit of background, auto generator on a call with credit unions and my wife just, uh, decided to take it upon herself to make that my background.

Steve’s Journey in Credit Unions

Doug English: Very nice. Steve, tell us about your journey in credit unions.
Steve O’Donnell: Oh, gosh. Uh, started as a teller back in 1997 actually for, uh, uh, we were Nevada Federal Credit Union at the time.
And, uh, back in 2010, uh, we changed our charter to a community charter and in a state charter and, uh, in our name to one Nevada. And, uh, from there I’ve kind of just, uh, moved through the ranks. Uh, uh, I’m, I’m the COO now. I’ve been here for 28 years. Uh, actually, uh, in that role a couple of months now. I’m a recovering CFO and it was at that time where I met Saroop, uh, through the CU two folks.
And, and one of the things that resonated for me, uh, going through listening to him and how [00:03:00] passionate he was about the AI movement and, uh, you know, doing a lot of reading, you, you couldn’t help but hear AI and everything you talked about at that time, and still at this time today. But it was, it was really interesting from an efficiency standpoint and understanding how credit unions are starting to think about relevancy more.
And when you, when you think about that, you gotta be maybe less leading edge and more bleeding edge in some of this stuff. And so that’s kind of where we went and started our conversations and started exploring all the possibilities. Uh, and, and I think it was even at that day when we first met, we had a conversation and it was like 20 ideas, uh, rolled out of that conversation.
And then a week later we had 20 more. And so obviously you gotta start small on some of these things and get those quick, easy wins. But, uh, I think the future of this is, uh, very exciting.

Credit Union 2.0 & Decision-Making

Doug English: Yeah. So let, let’s back up a little bit. So you started, you were already attending Credit Union 2.0 [00:04:00] events, right? That, that’s what you’re saying, Steve.
So you, you’re already attending those events and that’s a group of credit unions that I believe are kind of looking, leaning into technology, looking for what’s coming. What, what was that? A board initiative? Where did that, uh, uh, what, how did that get decided that that was something you’d be doing?
Steve O’Donnell: You know, it is something I kind of threw myself into.
Um, and, and this was, uh, through, you know, years of a relationship with, uh, Kirk Drake and Chris ote. And, uh, and, and Kirk had, uh, invited me one time. He said, you gotta be a part of this. Uh, ’cause you know, we, we were doing already some really cool things as a credit union. Uh, and a lot of that wasn’t, wasn’t so much a board initiative, but the board supported it.
Mm-hmm. And the reason that they supported it is because we had years of becoming more efficient. I mean, we, we started getting away from teller lines 20 years ago. Hmm. And so when we were looking at it, we were trying to be digital first way back then, [00:05:00] uh, and how we deployed, uh, deployed services and help our members be self-service.
So, uh, a lot of those things require us to think outside of the box.

From Idea to Strategy

Doug English: Yeah. So, so, um, you met Saroop, and that was a, a, a, a great connection, but was it specifically, did you come to the meeting, uh, thinking that the, uh, the, this particular implementation of AI was how you were gonna go about it? Or like, or did you, did the connection, uh, is where you sourced this particular, like, which we’re gonna, we’re gonna unpack in a minute, like how you guys, uh, used ai, but did you have a strategic process around which, um, implementation you would use for ai?
Steve O’Donnell: No, we didn’t. Uh, you know, I came into this looking at the agenda going, that’s gonna be one of the more important topics that we’re gonna cover during this two to three days that we were there. And, and I was really interested in it, but it wasn’t until Saru unpacked a lot of what the capabilities are [00:06:00] and, and gave us that rundown of what was really happening in that space that we started, uh, seeing some of the ideas.
And, and immediately the things that came to my mind is we, we had a big data initiative that we were working on. And so being able to connect ai, uh, even though in its infancy, I mean, there’s a lot of guardrails and things you gotta think about, um, but just, you know, thinking about where that could go and connecting the data and how that can empower our credit union to do things differently.
Um, you know, your mind just started spinning.

Early Generative AI Context

Doug English: Hey, Saroop. Can you do a mini version of that right now? Like what Steve heard that caused, uh, caused this direction to become what one Nevada did.
Saroop Bharwani: Yeah, I’ll share from my perspective, and that was about just over two years ago, which feels like a very long time.
Yeah. I was thriving in, in, in ai. Months. Weeks. It, it just feels like eons ago. And I remember, you know, Kirk and Chris invited me there, uh, in March, 2023 to [00:07:00] present about just what generative AI was and like I did. And I think that it just sparked a lot of conversations between someone who’s, you know, coming from the outside in and someone like Steve and you know, Gary and Joey, and the great people that we partnered with who have been in the credit union space for a while.
And that clash of like different perspectives and different mindsets. Actually is what, you know, Steve was talking about in terms of one idea led to 20 and led to 50. And like, you know, there’s so much you can do with this technology and, and you know, we’re at the early stages of a platform shift and in the early stages of a platform shift, it’s like everybody’s inventing stuff as we go.
The playbook is being created like right now as we talk, right? And so that’s so much fun. Like just being a builder and a creator. I love to work alongside like-minded folks that know industry really well to create impactful solutions. And it’s iterative and you just gotta find the right partners that are able to be iterative till you get to a point of success and then you can build on top of that.
And you know, Steve, you know Steve and the others, they’ve been perfect partners and early [00:08:00] adopters to really set the foundation and be trailblazers in their industry. And now, you know, the, the rest of the industry has. You know, has the benefit out of this because of the, the, the work we put in, in the early days.
So it’s super exciting to, to be where we are today, but it does feel like the last two years have really, really been like, you know, a lot of iteration and I’m talking about daily and, you know, it’s not only working with credit unions, but the models are improving every day. So keeping up has been, uh, quite challenging, but so much fun.

Core Use Case & “Response Quality”

Doug English: So what’s the, uh, use case? What’s the implementation? How, how did you talk me through that?
Saroop Bharwani: So I’ll, I’ll start and then Steve can talk a little bit about from an implementation standpoint, but what my team has always been really, really good at my, Tom, my co-founder and myself, and, and the team we put together is, is really processing data and, and just building models from scratch.
Started with classical models now LLMs and like, where we always sort of leaned into was knowledge. The, the fundamental first principles thinking is like in this AI wave, knowledge [00:09:00] is going to be that one thing. That never gets disrupted. In fact, you’re always gonna need more knowledge. People and members are always gonna have more questions.
So we really kind of like went down the rabbit hole of how do we take it upon ourselves to be the best at improving knowledge quality or you know, what I call improvement engineering, right? And so we really dug into that and, and you know, what we’ve been able to do over the past, like couple of years across the credit news we work with at Steve’s organization is really ingest policies, procedures, guides, basically the entire knowledge base from like.
Decentralized knowledge sources throughout the organization, consolidate that, make it accessible to MSRs on the front line, in branch and in the call center. And really sort of like just go through this improvement cycle of analyzing what works and doesn’t work on calls. Um, we also ingest a lot of anonymized call data to assess what works and doesn’t work.
And that feedback loop allows us to continually improve knowledge quality. And we have a metric called response quality that measures like the quality of your content [00:10:00] as it goes from like typically the average is 30 to 40% when a credit union starts with us and they get up to 80 to 90%, um, within, within weeks.
And, and those, that has a huge impact on, on call center performance. And that’s just one use case we’ve uncovered that we’ve been successful on across like Steve’s organization as well as other organizations as well.
Doug English: I wanna, I wanna understand that metric. What was that, 30 to 40% metric?
Saroop Bharwani: Yep. It’s called response quality.
And it’s essentially, the metric is the, your contents effectiveness in actually providing, um, you know, MSRs in this case with effective information at the point of interaction with a member, uh, across calls and branches. And what we find, there’s, there’s subtle nuances when we analyze the context of a conversation, a call, there’s little nuances where MSRs struggle to access information.
They may put a member on hold because they can’t find information in SharePoint, they have to call back office. The more people that call back office, there’s, you know, a, a bottleneck in the organization and all this stuff compounds. Whereas if you’re able to provide like, and [00:11:00] democratize access to information freely across the front line and measure the effectiveness over time, essentially you’re creating better member experience.
’cause there’s less wait time, there’s higher resolutions and less escalations and just the members feel better. And we’re able to measure all of this down to the very granular detail of what’s in the policy procedure guide. And that’s just created a lot of impact and continually improving on that is the foundation of where an organization needs to be in order to be what I call optimally efficient.
Doug English: And, and, and I think the metric you said was about double, he about doubled that, uh, that that metric.
Saroop Bharwani: Typically we, we see about 30 to 40% to 80 to 90% res results in about a five x reduction in back office calls, uh, significant, uh, member less members are on hold and it’s, it’s quite impactful in, in the get go, something you can achieve in weeks.

Planning, Board Buy-In & Metrics

Doug English: Oh, and I bet you there’s some other metrics that get, you know, new, new, new accounts conversions, additional deposits. So, Steve, let, let’s go over to you, Steve. So go back to the beginning of, uh, [00:12:00] of how you decided to implement and then what the rollout to process has been and, and any metrics you can talk about.
Steve O’Donnell: Yeah. So, you know, for us, uh, you know, when it started, when we, when SAR and I first met and we had a discussion and he kind of gave us a glimpse into what he was thinking and, and we thought that fit into a lot of different strategies for us. In fact, everything he was just talking about, you know, I’m just starting to think of other ideas that we might be able to do now.
Um, but, uh, you know what, the first thing that we did is we started thinking about a plan. Then I shared it with the management team here and, and just kind of talked through use cases and scenarios and what we could do with this. And then, uh, we actually had Kirk and, uh, Saroop, uh, attend our, uh, planning.
Session with our board, uh, at, at the, towards the end of the year as we started formulating our strategic objectives for the next year. And, uh, it was basically a situation of them giving a rundown to our board of what AI is [00:13:00] and what it can do. And, and they had a, a real, they were all blown away by it.
They were, uh, really impressed and they thought, this is, this made a lot of sense and it fit our model. And, and so from a metric standpoint and, and just using the use case where he is talking about calls, I, I think that’s a good one for us to look at because, uh, you know, I talked about 20 years ago we started shifting away from tellers and, and, and getting people to use that branch that’s in their pocket.
And when we’re doing that, we’re sitting at about 75% of our members utilizing that, uh, on a daily basis to access the credit union. The unintended consequence with that is they don’t come in your branches anymore. Mm-hmm. And so you don’t see ’em, you don’t know ’em, it’s harder to cross sell, but they do call into your contact center.
And so one of the things that we’ve always been big on is getting that. And, and this has been a, a, a, a, uh. A strategy for credit unions for a long time is a 360 degree view of your member. And that’s basically looking at all the data of your member and being able to, what’s the next best [00:14:00] product to sell?
Uh, what’s the loyalty look like? What do you need to do to, to push them in into another direction? Or if you’re seeing attrition, trends, et cetera. Uh, but what you don’t have is you don’t have sentiment. And so we may have a member that’s contributing to the cooperative and, and doing a lot of good things, but when they call in, they’re just irritated all the time.
So that was a piece of data that wasn’t, uh, a part of our scope. And now that’s something that could be part of that scope and, and will use that to look at it. And then we need to have different conversations. Uh, we, we believe, uh, at one Nevada that, uh, we need to be showing those members love all the time.
And, and that’s something that we’re really focused on and I think this will help us get there.
Doug English: Have you changed the credit union’s tagline to love all the time? ’cause I think that’s a great thing.
Steve O’Donnell: Well, it was our, our one focus is you. So maybe we’ll work on that.
Doug English: I like it.

Current Performance & Sentiment

Saroop Bharwani: And, uh, Doug, uh, if I could add a couple of stats in there on when Nevada, they’re amongst the highest response quality across all [00:15:00] credit unions.
They’re at 92%, so they’ve been maintaining their documents incredibly well, and I’m looking at their conversational data, their calls, and, you know, we see resolution rates, uh, going up. We see, uh, sentiment getting better. We see handle times, uh, reducing across a number of categories. Now of course, there’s always other categories to improve, but we can tell across like, you know, any type of loan across payments, across disputes, um, where the issues are and, and really get context on how to resolve that.
And, you know, working closely with, you know, Steve’s great team in terms of chipping away at how to get to optimal efficiency is now a data-driven exercise. It’s not like, you know, in invisible to anyone. It’s like all the data is all of a sudden in front of you and you’re able to actually consolidate thinking in terms of how to solve these problems and chip away to, uh, to optimal efficiency.
Steve O’Donnell: You know, I’d, I’d add to that too in terms of, um, you know, one of the focus that we have here is, uh, being able to upskill our staff and let them have more time to engage with the member on deeper relationships. [00:16:00] And doing these things with AI in the front of this is allowing us, because resolution is happening at, at a quicker array.
Uh, so it allows our staff to do other things and, uh, gives us the opportunity to upskill. And so it’s that, it’s that adage where everybody’s talking about AI’s gonna replace things. Well, we’re, it’s gonna augment things for us, and we’re working on doing that.
Doug English: Yeah. That’s really interesting. Is there any examples, like you can decide of, of what that upscaling is?
Steve O’Donnell: Yeah. Uh, one of the biggest examples that we have is, uh, you know, uh, a few years ago, the biggest metric for us from a contact center perspective was how quickly you can get that member off the phone and then get to the next call. So the service levels were really strong and we didn’t have phone, uh, calls dropping off.
Well, now that has completely changed, and our average call is, is increased by two or three times, and that’s because we’re engaging with those members that really need things answered, and we’re not rushing ’em off. And so we’re creating better loyalty. We’ve seen our attrition [00:17:00] numbers drop as a result.
And so I think we’re seeing a lot of positive things from that, that we’re still learning, you know, every, every day we’re gonna see different things that, uh, are, are, uh, results of these impacts and changes that we’re doing.
Saroop Bharwani: And that’s something we learned too. Uh, Doug, just sorry to interrupt, but we learned that too.
Initially, our mentality was you want to get average handle time as low as possible across everything. And then when talking to Steve and his team, they’re like, wait a minute, no, actually, we want to actually spend more time engaging them. And we analyze the data, like, you know, sentiment is higher on a lot of those calls.
And even like, really what I call elegant cross sells are happening where, you know, it’s actually a product that they need and want that’s tangential to the conversation they’re having. So it doesn’t seem like a, you know, a sleazy salesman. It seems like an actual, like, you know, advisor helping a member.
Um, and it’s really, really incredible to see that I, I learn new things from, from Steve and his team every day.

New Practices: Outbound & Proactive Service

Steve O’Donnell: I’ve got one more use case that, uh, has been really positive for us too. And it’s [00:18:00] looking at the contact center, not just as an inbound call, but we are now doing outbound calls because they have time to do that.
And so now we’re calling members and you know, it’s a, like the pandemic was a, a good example. We would reach out to ’em and say, Hey, how are you doing? Um, but that was a concerted effort. Now we have the time and cap capacity to be able to do that and have different discussions with members. Uh, especially if, if problems arise, we’re, we’re getting in front of it and, uh, they feel like we care about ’em ’cause we do.
Doug English: So let me, I, I wanna open that up a little bit. So what, uh, is the staffing the same, the same number of, of staff and you’re spending more time on the calls, but yet you’re handling more calls? Uh, or, or like, how is it that there’s extra time if you’re, if the calls are longer and you got the same amount of staff?
Steve O’Donnell: I know, it’s amazing, right? Yeah. It doesn’t add up. Yeah. Um, so what, so what’s happened is, uh, you know, [00:19:00] we’ve, we’ve got, um, a couple of different tools we’re using, um, uh, chatbots and, uh, we, we’ve got a tool posh that’s in front that senseo then is on the back end of that. And so all of those things combined have positioned our calls.
It used to be 28,000 a month that we would get. Now we’re down to 16 to 17,000 because they’re being handled. And so the ones that are important get to the agent. The agent deals with those, and that frees up the time,
Doug English: so the, then the members are getting their question answered by other systems. Right.
And it’s only the calls that need that human interaction that are making it through to your call center. Mm-hmm. Is that, is that the right way to summarize that?
Steve O’Donnell: Yep. And then those calls that are getting through, uh, you know, we’re learning so much about ’em through the Senso platform.

Beyond the Contact Center: Web & GEO

Doug English: Very interesting. So, um, that’s [00:20:00] been now have, have you expanded the use case beyond the, uh, the, the, you know, the, the source of the, the best training, the best knowledge for your, uh, call center representatives? Is that li is, is that grown into other areas of the credit union? I’ll let you, uh, comment on that.
Saroop Bharwani: Yeah, no, I would love to get, I mean, this is very recent, but, um, you know, again, like working with a very collaborative group of credit unions, you learn about new use cases and new ideas, and one that emerged sort of from the dust, uh, was marketers asking us, Hey, you’re improving knowledge for your policy, the credit use policies, procedures, and guides, um, can you do it on our website?
And so we started to go, go through a, an analysis about, like six, six months ago on, Hey, what, what if we can identify gaps on websites using the same engine, um, and really sort of recommend to co uh, marketers on how to fill those gaps. Well, you know, uh, what really kind of like, and this is completely serendipitous, what started to become really important.
Is, you know, the [00:21:00] transition of behavior from searching on Google for 10 blue links to searching on chat, GPT, perplexity, Claude Gemini, meta AI now, and the many other AI platforms, that transition is happening very, very quickly and now all of a sudden, instead of optimizing your answers for keywords, uh, search engine optimization, which is allocated budget that every marketer in the world is, has been spending time on for the past two decades.
Instead, you need to optimize for generative engine optimization and optimizing for the optimal answer on the AI search platforms. So what we’ve been able to do is leverage our content, uh, improvement engine for the improvement of your web content to be more visible and relevant on the major AI search platforms.
And we’ve just rolled that out. Um, we’ve set a baseline across about 200 credit unions to date and you know, we’re adding 50 or 60 new ones a week. Um, you know, Steve, uh, and his team have their visibility scores on how they rank [00:22:00] within their state, uh, within Nevada, and it’s hyper localized. We can get down to the community level in terms of how visible you are in your community.
And the beautiful thing is we’re not only ranking each credit union within their community, we’re ranking the movement as a whole. So this is how credit unions as a whole are ranked relative to other financial institutions like banks or mortgage providers, et cetera.
Doug English: Is that something you can show us and talk through while we’re on the podcast together?
Is that, uh, content that’s appropriate for this space?
Saroop Bharwani: Yeah, absolutely. A hundred percent. Um, I’d like can, I mean, go ahead. I mean, Steve, if you’re good with it, I could share your, uh, I could share your insights in terms of how you’re ranking, uh, on, on chat GBT.
Steve O’Donnell: Yeah, actually that’s a good one. Yeah. Awesome.

GEO Demo & One Nevada’s Visibility

Saroop Bharwani: So what you’re seeing right now, I mean, uh, g do sensor.ai, front slash credit dash unions. And essentially what this is, is it’s, think of it as a network.
Uh, it’s a network of credit unions and we’ve ranked, uh, but [00:23:00] 200 credit unions right now on how visible they are on chat GPT. Okay? We’ve just started there. You see the leaderboard over here. Top ranked credit unions, um, that have been, you know, that have a GEO score. And again, generative engine optimization is what is your share of voice and what percent of the answer includes your credit union in your local locality.
So I’ll switch over to what the post dashboard looks like. So what you see here is one Nevada Credit Union. And they’ve been mentioned out of 10 questions we’ve asked that are trending in the credit union space in Nevada. Uh, we’ve, they’ve mentioned nine outta 10 times. Their visibility ranking is about four.
And their share of voice, the percent of, uh, the answer that includes one Nevada is 7.6% relative to 7.2%, which is the average in the industry. So they’re actually doing better than the average in Nevada, right? Which is a very good thing. Now, there’s always room for improvement. We want to get, uh, share of voice up [00:24:00] to 10%, which is a good, you know, percentage, 10%, um, visibility ranking.
We want, obviously want to get to number one and we wanna get them to mention in all, in, in all answers. Um, so here’s an example. It’s a college student in Nevada, which credit union have the top no fee student check in with nationwide ATM access in this answer. Uh, and this is just on chat, GBT, uh, they came up 10th when Nevada came up.
10th. And, uh, their share of voice was 10.2%. So 10.2% of this answer down to the character level, uh, includes uh, one Nevada, uh, but they were 10th. So there’s a lot of room to improve their content on their website to get them ranked higher.
Doug English: Yeah. Now that only looks like it compared to other credit unions.
I, I imagine if you let that compare to banks and fintechs, uh, would it be interesting to see what that look like?
Saroop Bharwani: Yeah, exactly. I think some of these answers would compare you to banks and fintechs. So I think we did this exclusive to credit unions, but we’re starting to open it up to the broader base because again, it’s important that you’re not just ranking against [00:25:00] your peers in the credit union space, but like, you know, you’re ranking higher against the broader financial services network.
Steve O’Donnell: And then there’s an example number seven there. It said Ally Bank. Mm-hmm. Credit union. There you go. Technically a bank, but has a digital option.
Saroop Bharwani: Yeah. Yep. You know what? One thing I will say is that one thing we notice here is that like. When the GPT generates a result, a lot of the time the rates are wrong, the terms are wrong, and there’s just a lot of room for improvement because we have, like, you know, Steve’s first party data really making the answers more relevant, um, is something that, you know, senseo specializes in.
And really what questions to ask. ’cause we have the call data. It’s like we can ask the most trending questions based on real voice of member in Nevada, right? And really that’s the key for what makes essential platform special and unique, is that we’re able to improve based on real questions and we’re able to improve based on real first party data that’s unstructured that all of a sudden becomes relevant [00:26:00] in these answers.
So that, that’s really what we’re excited about, um, for the future.

Next Metrics & Goals

Doug English: Keeping, getting better and better and better, getting better training for the member service reps, resulting in better, uh, service to the member. And hopefully Steve might even have some data on additional cross-selling success or new accounts.
Have you seen any pickup in that area, Steve?
Steve O’Donnell: Uh, I think that’s still in its infancy stages as we, we try to connect some data points, but that is the goal.
Doug English: Yeah. I, I, I would think so. Now, Saroop, does this, uh, uh, data on optimizing for, uh, for the, uh, the, the models did what, what’s like, you see where you’re ranked and then what do you do about it?
Saroop Bharwani: Yeah, so that’s the beauty of it, because we have so much first party data from one Nevada internally, their teams are already using our improvement engine internally to understand what questions can and can’t be answered. And then it’s just a exercise of [00:27:00] structuring that information the right way. It into their cms, so that robot txt captures it.
There’s some configuration work we do there, and it’s just continuing to reran, uh, one Nevada across these questions on a weekly, monthly basis until they get to a high improvement RA ranking. The point is, it’s in our control.
Doug English: Yeah. And it’s a whole, it is a, it’s a mental shift for the marketers to go from optimizing for SEO to optimizing for an AI model.
Uh, it, it, it’s one that I have not had a discussion on before, so thank you for bringing it up. Uh, I think there’s a whole lot more to that. Um, backing up for, uh, for example, uh, on, from what you’ve done with Steve and the team at One Nevada to, can talk to us about, uh, uh, any different use cases, especially really successful ones across the credit union movement.
How many credit unions are you working with right now? Sar Ruper about.
Saroop Bharwani: About a dozen. Um, and that’s, [00:28:00] that’s growing quite quickly now. Yeah. We’ve been incubating the product with the, with about, we incubated it with about seven credit unions in the first year, and now we’re really hitting our stride, uh, with the solution now that it’s stable and it’s, it’s functioning and have, you know, credit music, Steve, uh, Steve’s to thank on on really being that early adopter for us.
Doug English: Mm-hmm. And, and are the use cases, this, this, the, the, the, uh, training and knowledge bank, uh, each time? Or are there other use cases?
Saroop Bharwani: We’ve, we’ve gone deep with knowledge improvement because that is a rabbit hole, infinite rabbit hole. Right. And, and, you know, being focused is critical and we feel that knowledge is the center of a lot of like, really, really impactful use cases.
We’ve talked about two call center optimization. We’ve talked about visibility in on the ENT web, uh, visibility on your website. Um, I, you know, I’m seeing a lot of other use cases out there, right? There’s, you know, great solutions on the unified communication side. You know, Steve talked about Posh and there’s other ones there that are really kind of like focused more on the, you know.
Actually in the engagement of the member. Uh, and, you know, [00:29:00] we’re, we’re the, we’re a feed into that, but really that engagement side requires a, a lot of attention there. That’s, that’s a very obvious one there, there’s a lot of fraud use cases out there, like capturing fraud, reducing false positives, right?
Like,those are like very, very impactful. And I would say that as impactful as AI is for good things, detecting fraud, there’s also more fraud happening. So these tools are making it easy for like malicious actors to, you know, really kind of like get creative with regards to how to like, you know, you know, voice phish a member and, you know, access their account and various other things.
I think I remember in the board meeting with Steve, I think I mimicked, you know, your CEO’s voice in front of everyone and like, you know, everyone was kind of stunned, right? But it really made it real in terms of like the threats that are going on as well, uh, which is, uh, you know, essential in the, in the financial services industry.
We always wanna do the fun stuff, but like, there’s the protection side as well.

Getting Started & Community

Doug English: Mm-hmm. Very much. Well, as we wrap up our [00:30:00] discussion, if you can think, uh, for a listener, uh, on this, uh, someone listening to this podcast that might not yet have decided how to implement, where to, to go first, uh, what to use to guide, uh, their AI strategy.
Um, if, if you can comment around what you, you might, uh, suggest that they do, uh, Steve, I, I particularly love it for you to talk a little bit about the Credit Union 2.0, like what, what you’ve gotten from that and what, what your thoughts are around there. And then, uh, uh, again, just any general thoughts around how should a credit union start if they haven’t yet done so.
Steve O’Donnell: Yeah, so starting with the, um, the 2.0, you know, for, for us, it’s just been a really good spot for collaboration. I mean, meeting with a lot of different credit unions and, uh, fintechs that are thinking about the same things and, and a lot of us have the same problems we’re trying to solve, [00:31:00] and we’re all working to that end.
And so the collaboration has been really interesting. Uh, even with the fintechs, I mean, I, I’ve, uh, had many opportunities to get a couple of fintechs together in the same room and say, this is what we’re dealing with. It looks like each one of you can kind of get together and solve this. And, and, and I’ve seen some of that happen.
So it’s been very interesting from that standpoint. And as far as, uh, you know, getting a credit union to start and, and, and take the first step on this, I, can we ask that chat, GPT, is it still listening?
Doug English: I muted it. It was just a too eager to talk to us, but we can, we can ask it. But let’s, let’s give Saroop a chance first.
Saroop Bharwani: I don’t see you too. Kirk and Chris. I mean, great, great advocates for the credit union movement and, and just technology in general. And, you know, Kirk’s all in with ai. Um, so is Chris. Um, I would say also Lamont Black who we, you know, saw at Edge last week is really like spreading some really good thought leadership.
And I know that. Um, you know, it’d be great if he came on your pod. [00:32:00] Um, he coined a term last week called Artificial Business Intelligence. And really, I mean, describing like in words what Senseo has built, uh, from the ground up. You know, being able to kind of really access unstructured data in a very flexible way to generate content, extract insights is essentially what the basis of the central platform does.
But having that thought leadership from someone like Lamont has been like really, really impactful from the top down. So, you know, he’s, he’s a great resource and I would recommend, you know, reading his material and artificial business intelligence is the, is the term. Um, we’ve developed an AI sandbox.
I’ll send you the link right here at sensor.ai/sandbox for credit unions to just want, it’s like a playground. You want to get started and really start to kind of like, get your feet wet with the tools. You know, literally just, we’ll give you access and just get a glimpse, right. A lot of the time it’s like.
Credit unions are unaware of what’s possible, and if you kind of like, you know, make it possible, like, and show them possible, like very easy, they’ll start to unlock like [00:33:00] superpowers in their heads and like, you know, all of a sudden they’ll dive in. And then lastly, I like, you know, I spend most of my time in Silicon Valley and, you know, I, I do regular hackathons, so I’m seeing what’s emerging from the dust with like the development community there and really bringing that knowledge back to IT.
Teams within the enterprise that are struggling with things like consolidation of data and engaging those IT leaders like Jesse at, at One Nevada and many others on calls on how to kind of get through these barriers is another thing that I, I actively do on the education.
Doug English: Well thank you for that. We’ll, we’ll include those notes, uh, in the podcast and, uh, we’ll reach out to Lamont Black and see if we can get his artificial business intelligence, uh, uh, as a part of one of the episodes coming soon.
So, uh, Saroop and Steve, thank you so much for finding the time, uh, to help, uh, us capture the leading ideas for the credit union movement in ai. Uh, I love what you’re doing for your membership, uh, for the credit union movement as a whole. Uh, and, uh, don’t slow down ’cause it’s only gonna [00:34:00] get faster. That’s right.
Yep, exactly. Thank you guys.