FintechTalks · 2026-05-04 · 15 min
Key moments - from our scoring
Substance score
55 / 100
Five dimensions, 20 points each
Avre AI tackles one of fintech's most labor-intensive problems: the $300 billion global compliance industry built on manual review processes. Rhim Shah explains how his company replicates human decision-making in BSA/AML workflows, sanction screening, and transaction monitoring using agentic AI with robust guardrails. His path to founding came from leading financial crime at Revolute, where he witnessed tens of thousands of daily reviews conducted by teams of 500+ people - a clear signal that AI could add massive value once models matured. Unlike many AI startups claiming quick wins, Shah emphasizes the need for 99.99% accuracy in compliance, which is why Avre invests heavily in research-level product development rather than simply wrapping ChatGPT. The conversation covers go-to-market strategy (40-50% of resources on research and product, the rest on sales and distribution), the critical role of relationship-based selling with financial institutions and core banking providers, and how FinTech Meetup's authentic format has proven more productive than larger conferences for booking qualified meetings.
Avre AI deploys agentic AI to automate financial crime compliance operations including BSA/AML, sanction screening, transaction monitoring, and investigations that are currently manual and error-prone, serving banks and fintechs. The company replicates human decision-making processes using AI with robust guardrails to achieve 99.99% accuracy.
It is a $300 billion global industry, with $60 billion in the US alone, built on hiring humans to perform manual and repetitive compliance review workflows.
After seeing the massive scale of manual compliance work at Revolute - tens of thousands of daily reviews by 500+ person teams - and recognizing the potential trajectory of AI models, Shah decided to found a company focused on solving the specific problem domain he deeply understood from his experience leading the financial crime product team.
Avre AI uses warm introductions from advisors (ex-CROs and risk officers from major banks), backtests alerts through their system to show actual results, and emphasizes deep research-level robustness rather than simply applying ChatGPT, combined with strong founder relationships to build institutional trust.
He allocates 40-50% of resources to research and product development to maintain accuracy and robustness, with the remainder split between go-to-market and sales expansion, plus additional investment in branding and conference presence to establish market awareness.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode contains some substantive information about the BSA/AML compliance market size ($300B globally, $60B in US) and the founder's experience at Revolute, but relies heavily on soft advice about relationships, trust, and founder mindset. There is minimal technical detail about how the AI actually works, what specific accuracy improvements were achieved, or concrete metrics on customer outcomes. Much of the conversation drifts into generic founder wisdom rather than actionable insights for B2B operators.
there's a $300 billion sort of industry built around hiring you know humans to operate these very manual repetitive workflows. In the US alone, that's that's 60 billion
even for a fintech like Revolute, there were still, you know, tens of thousands of reviews done daily by a team of you know, four or five hundred
The framing of deploying agentic AI to reduce false positives in AML compliance is sensible but not particularly novel - this is a well-trodden application area. The emphasis on relationships and trust-building as a go-to-market lever is standard startup playbook. The discussion of accuracy requirements (99.99%) is relevant but presented as straightforward necessity rather than generating fresh strategic insights. No contrarian claims or first-principles rethinking are evident.
agentic AI to automate financial crime compliance operations, BSA AML within banks and financial institutions
if the bank in an FI can't trust the person, how can you trust the product, right?
Rim Shah is a legitimate founder with relevant hands-on experience: ML background, time at Google, and lead of the financial crime product team at Revolute. He has direct insight into the problem space and has built products in this domain. However, he is still early-stage (under 2 years at Avre AI) and is not a seasoned operator with proven scale-up experience or a track record of building category-defining companies. He is solidly practitioner-level but not the caliber of a veteran battle-tested executive.
So my background's in machine learning, uh so I have quite a technical background
spent a bit of time at Google, um, and then after that ended up going to Revolutes, yeah. Um, where I eventually ended up leading the financial crime product team
While the episode names Revolute, Google, and Avre AI, and cites the $300B and $60B market figures, it lacks concrete operational metrics: no customer names (beyond 'large institutions'), no specific accuracy/false positive reduction numbers, no timeline data on customer implementation, and no revenue or growth metrics. The discussion of the 99.99% accuracy requirement is stated but not evidenced with examples of what happens when that threshold is breached or what competitors achieve. Advisors are named generically (ex-CRO of Wells Fargo, head of risk at US Bank) without specifics on their contributions.
there's a $300 billion sort of industry
we're working with some very large institutions
The host asks open-ended questions and does follow up on some responses (e.g., diving into how he convinces banks, probing the chip allocation question). However, the conversation often meanders without sharp follow-ups on key claims - e.g., when Rim mentions 'large institutions' and 'core banking partnerships,' the host doesn't press for names, deal sizes, or status. There is no pushback or productive challenge on claims; the dynamic is warm and affirmative rather than investigative. The host's questions are friendly but lack the incisiveness needed to extract concrete evidence.
how do you convince them of that?
if you had a hundred chips, if you will, like how many, how much of how many of those chips would be on like technology
Computed from the transcript - who did the talking, and the words that came up most.
In this episode of FintechTalks, host Sanjib Kalita sits down with Rhim Shah, Co-Founder and CEO of Arva AI, live at Fintech Meetup. Arva AI is using agentic AI to automate financial crime compliance operations across banks and financial institutions, including BSA/AML workflows, sanctions screening, transaction monitoring, onboarding reviews, and investigations. Rhim shares how his background in machine learning and experience leading financial crime product at Revolut shaped the idea for Arva AI, why compliance operations are still heavily manual, and how AI can help reduce false positives while maintaining the accuracy, explainability, and trust required in highly regulated environments.
Transcribed and scored by The B2B Podcast Index.
Welcome to another episode of FinTech Talks. I'm Sanjeep Khalida and I am lucky to be here today at FinTech Meetup with Rim. Rim, would you mind please introducing yourself? Yeah, definitely.
Thanks so much, Sanjeeb. So my name's Rim. I'm the co-founder and CEO of a company called Avre AI. And we deploy Agentic AI to automate financial crime compliance operations, BSA AML within banks and financial institutions.
And uh how how long have you been doing this? So it's been a couple of years now. Um we started our journey in San Francisco and now we were based in New York. Yeah.
Um, but it's been been just under two years. So what what's uh uh could you get dive a little bit more into what what RIM actually I mean uh Arva actually does? Yeah, definitely. Um so if you think about your your average or really any bank or fintech, um, you know, they have a large BSA email program or even a smaller one, yeah.
Um but the entire program is is ridden with you know false positives, right? Yeah. Um when it comes to things like sanction screening alerts, transaction monitoring, email reviews, investigations, onboarding, all of this is largely driven by humans right now. Um in fact, if you think globally, there's a $300 billion sort of industry built around hiring you know humans to operate these very manual repetitive workflows.
In the US alone, that's that's 60 billion. So it's a massive, massive sort of you know space. Um so our goal is really to understand what the human has to do, replicate that same process using AI, deploy it across that bank or a fintech, uh, and process those reviews in the same manner. Um so fundamentally it's about understanding what the human does at alert generation and trying to process it in the same way using agentic AI with really robust guardrails in place.
Wow. And and and uh well first of all, I would never would have guessed that that market was that large. And and and and uh so how how did you wind up getting into this? Yeah, that's a fantastic question.
So uh so my background's in machine learning, uh so I have quite a technical background. Yeah, um, always loved building products, spent a bit of time at Google, um, and then after that ended up going to Revolutes, yeah. Um, where I eventually ended up leading the financial crime product team within RevLute business. Yeah.
Um so I built up a lot of the different back office ticketing, transaction monitoring onboarding platforms for our team back at the time, yeah. As well as a lot of customer-facing flows too. Yeah. And I think it gave me a really good understanding about the pain points and how manually intensive it was.
Even for a fintech like Revolute, there were still, you know, tens of thousands of reviews done daily by a team of you know, four or five hundred. Um, and I think it was around the time when Chat GPT first kind of you know became mainstream, yeah. And the realization kind of struck that actually, yeah, this is a fantastic application where models might not be good enough right now, but there definitely could be that trajectory that at some point in time this is a fantastic application.
And and uh it's there's so many potential questions to go from there, but uh how like it's it's it's it's sort of um the space is obviously so fast moving, and um it's it's sort of like what like how was there a point where you at Revolute you're like I just need to make something and and like sort of I guess like late make a leap, if you will. Yeah, it's uh another fantastic question. I I think actually when I left Revolute, it wasn't fast moving, right? So like the AI space was still very nascent.
Yeah, um, there wasn't that common understanding of actually this could be really impressive, right? I think you know you would try to add GPT and it would still make a bunch of mistakes and give you really, really horrible, interesting results. Um so at that point in time, AI wasn't moving as quick as it is today, yeah, right. So I think the the insights I had back then were were a lot more focused around actually where could this go?
What is the potential, not what is happening right now. Um, and I think you know, a big part of it was also just the the want to be a founder as well, right? So it was like, you know, I want to be a founder, yeah, and I I know there's also some pro problems to solve out there. Let me look close to home, right?
Let me look close to home where the problems I solve day to day, how can we deploy AI or some use case to actually solve things that I understand? Also get going back to that point, like so. I I actually met many years ago when I was in college, I I used to be in AI, I used to do research in AI. I um and I got published in uh IEEE uh biomedical engineering and about by developing artificial neural networks to classify electrocardiograms.
Wow. And back then like there the limitation was was hardware and and that now there's limitations on hardware and software, and like how how do you sort of m try to envision where things will be? And it's like it this is a good time to to make that leap. Yeah, yeah.
I think I think probably the best way of of honesty is actually how we approach how we build internally right now, right? So if you think about our problem space, it's not an area where you can be right, you know, 95% of the time, you've got to be right 99.99% of the time, yeah, right. The the margin for error error is so small and so slim, yeah, um, that accuracy and robustness is top of mind.
Yeah. You know, this is very much true for fintechs, but even putting fintechs aside banks, you've got to be really on it, right? Yeah, yeah. Um, and so I think for us, you know, our core thesis is around how can we build the most robust, explainable product.
Yeah. And that's actually why we get a layer deeper into the research side as well. We're not just application layer. We don't just take, say, your chat GPT and then put that in a product.
Yeah, we focus really deeply on the research side as well. Yeah. Um, and so I think for us it's just about internally within our own research teams, forget sort of you know, the the research labs outside, how can we push the frontier ourselves, right? Yeah.
How can we constantly improve um the accuracy and robustness? And and um the the the how like I so as a founder of this fast moving company with I'm sure like very talented people, like how what what what's um what you know that like what as as a leader, like you you know you you get called upon like some technical knowledge, some you know, vision, like what what what what like what what do you what does it draw from you that you enjoy doing more so than you ever realized you you would have enjoyed?
Yeah, well, I I think for me everything that I enjoy is our relationships, right? So whether that's skating the team and building relationships internally, whether that's uh you know advisors and investors, whether that's you know prospects and customers, um, I think a lot of it is relationship driven, yeah. Um especially in this this industry and space, because if uh if the bank in an FI can't trust the person, how can you trust the product, right? Yeah.
Um so I think a big part of it is actually playing on those those assets of mine, I would say like what I enjoy anyway. Yeah. Um, and I think the most surprising thing is actually how important that is for hiring, in fact. Some of the most important hires we've made you know internally within the company have been ones where I've built relationships over several months.
Yeah. Um, even our even our chief revenue officer, uh, for example, who recently came aboard, is a fantastic example of someone who's incredibly talented. Yeah. And she took the leap of faith to join, you know, a company at that early stage because of the nature of the relationship.
Um so I think that's probably the most surprising thing. I don't know if that directly else is. Yeah, yeah, no, that that doesn't, and and it's like I'm not surprised because like when we met at FinTech Meetup, like I'd met like hundreds of people and you you you definitely like stuck out. And I was like, this is someone who I'd like to keep in touch with, and and I so so like when you're trying to talk to banks about this where you you have to be 99.
99% accurate, like how do you convince them of that? Yeah, I mean it's true because there's a lot of AI noise in the market as well, right? I think a lot of folks are coming to banks and fintechs and and claiming that they can do this and do that, yeah. Right.
A lot of them have been burned by running POCs and they just don't go to plan, or they can go to full like fully to production, they get sort of their issues come up, right? Yeah. I think part of it is the relationships aspect of it. Yeah, right?
It's making sure that even if you don't necessarily know the buyers and the stakeholders within their bank, yeah, you know, we have many, many advice that help support us by chatting to these folks. Well, there's sort of the warm introductional connection that that's present, right? Yeah. So I think that warm relationship really helps because then they know someone they trust already, and that that person already works with us, etc.
Yeah. Um, and the second part is just creating success through the VOC, yeah. Right, running all the different alerts through a back test and showing the results as they are. And and and uh so you you know, we're obviously at FinTech Meetup again, and and so like what what's where have you like from when we first met, what's what what are sort what what's sort of the the advancement in both you as Rim as well as Arva.
ai? Yeah, definitely. I think uh you know when we last met a year a year ago, yeah, companies are very different, very different shape and size, right? I think the challenges are different.
Back then it was um you know a much smaller team, we were still sort of developing products and working with sort of our early customers. Now we're working with some very large institutions, um, you know, exploring partnerships with you know, kind of you know, core banking providers and things like that. Yeah, um, so I think just the size and scale of the problem has has changed. Yeah, I think also if you look at the market, it's really matured and warmed up.
Yeah, where a year ago folks were kind of dipping their toes in the water when it came to AI and now they're fully engaging and wanting to actually deeply explore it. Yeah, that's definitely part of it. I think if you think about me personally or internally, I think it's it's a much sort of greater understanding around okay, now my goal isn't to be that I see who could do everything where I'd be everywhere now. How do I actually grow the company?
Yeah, how do I maintain leadership with the next layer down? How do I hire the best? Yeah. And so I just think from a challenger prompt perspective, that is what sort of has changed.
Yeah. Um, but the the market has also shifted. So I think a lot of things have shifted around us in the past year. And and and and so what what what what are you going to be doing this year?
And like what um believe you'll be speaking and on a panel, right? Yeah, exactly. So so this week, yeah, in a couple of days, I'll be speaking on the the AI risk and reality panel with you know so far wex and mercury, which will be really exciting. Yeah.
Um, you know, this this year, I mean a lot of our panel speaks to what we're doing this year, right? So the theme around actually how do we continue to invest in our product, how do we continue to deploy it across our customer base, um, and actually more deeply invest in relationships with some of these core banking providers and distribution partners that is now a bigger bet when it comes to banking, right? Um, so we have some really stuff, some exciting stuff planned around this.
Um, so a lot of a lot of stuff to come. Yeah. Definitely stay tuned. Um, but it's really just a continuation of that.
Build the best product and sell to the best companies. And and uh so in in in to achieve those goals, like you know, let's say if you had a hundred chips, if you will, like how many, how much of how many of those chips would be on like technology, how many would be on like the team, how many would be on let's say distribution? Like what how do you think about that? Yeah, that's such a good question.
Um, I think you know, a good proportion of them, at least 40 to 50, wouldn't be specifically on the research and product side. Um again, that that whole thesis around being right 99.99% of the time is core to how the product works and how we go to market. Um, I think the remainder will be in our expansion in sort of on your go-to-market and sales hub.
Yeah. Um and expanding distribution when it comes to that. Yeah. Um, and I think probably the second aspect would be um sort of more on the branding side, right?
So you know, a good proportion around conferences very much like this. How do I get our name out there? How do we meet new folks? Um, how do we invest in in ensuring that our name is is full forefront for what we're doing best at?
And in the and the it's uh sounds like you you've been making a lot of progress. How has FinTech Meetup been able to help you in in that process? Yeah, I mean, even the one day in has been phenomenal, right? I think the number of calls and meetings have been booked has has far dwarfed, you know, in one day, you know, any conference in three or four days, actually.
Um so I think it's incredibly productive to see that the format is really retained. It's it's kind of the core, I guess, the um authenticity of it. Yeah, right. The the meetings isn't something that hasn't gone away, which is really nice to see.
Yeah. And I think you often don't see that with other conferences, yeah. Especially the ones on larger scales. Yeah.
Um, so so incredibly excited for the next couple of days to see how that how that expands. Some really exciting meetings planned. Um looking forward to all the different talks that are happening as well, seeing what else is going on in the market. Yeah.
Um, so it's uh it's an exciting time to be in Vegas. It it it is a great time to be in Vegas. And and and uh it's uh the the thing is also like you you're out having fun, but you're talking about like you know, fraud rates or something. Yeah, some things that you normally don't are able to talk to general pop, you know, folks about.
It's uh it's really nice. Yeah, exactly. And and and and uh what are have have there been any kind like what what types of uh any interesting conversations that you had last year that helped you shape where where you've brought the company now? Yeah, yeah, definitely.
I mean, I think the the quality of folks that I was speaking to last year was similar to this year. I think this year we've done more. But definitely last year it was a similar scope scope of folks, and they've really helped us to understand the ICP around fintechs and banking, really understand what they're looking for, what they like, what they care about. Yeah.
Um so a lot of those conversations you know went in the right direction. So, you know, there's definitely there's definitely good ROI on on sort of attending last year. Yeah, from more of a direct short-term perspective, but from a longer term perspective, a lot of those insights, those conversations have helped inform a strategy overall longer term as well. And and um how about like as a founder, it's like one of the things it's is is it's like it is um it can feel very lonely, but but at the same time, like when you talk to other founders, you d there's a bit of like I I I don't know, like a bit of like it's almost like therapy, right?
And and and and and um like how Yeah what what sort of um have there been parts of you that like have like that that you are like so grateful that you've taken this leap for that wouldn't have been exposed if you hadn't taken that leaf? Yeah, yeah. I mean I think taking that leap has just shown a completely different perspective on everything, really. Yeah, yeah.
Uh you know, aside from kind of the more commercial and and business world, right? And also the entire of the the the I mean I'm from London from the UK, right? Yeah, that's where I started my journey. Um, but I think you know, leaping into the founding journey, you know, starting off from Y Combinator has really started uh really kicked off things in the US, yeah.
Right. So like getting good exposure to that, understanding how the entire ecosystem works for the past like several years, and also building up an incredible advisor base, right? So we're advised by folks at the ex CRO of uh the chief risk officer of like Wells Fargo, right? We're advised by the head of bottle risk management at US Bank, right?
Yeah. So I think taking that leap has given us exposure to to those kind of personalities and individuals who, yes, they've been great for the business, but also on a personal level, it's it's it's it's incredible to see such experienced folks, you know, engaging with with someone like me. So that's that's lovely. That's lovely.
And and and uh uh once again, like I'm uh I'm really glad that we had a chance to meet last year and uh grateful that we were able to have another conversation today. Uh Ted, thank you for joining us for it. Thank you so much to G. Really appreciate that.
Thank you. Take care.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
Why your research needs a “thinking cave” with Sarah KlingThe Curiosity Current: A Market Research Podcast · on Agentic AI89 / 100
Decision Logic: The Difference Between an Answer and a DecisionThe AI Forecast · on Agentic AI87 / 100
KYA Won't Always Protect You. The Real Risk Is the Swarm!Fintech Conversations & Insights with Efi Pylarinou · on Agentic AI86 / 100
Agentic AI in Sales: What Business Leaders Need to KnowScaling with AI · on Agentic AI86 / 100
95% of Emissions, 14% of the Capital: The Bond Market Mispricing in Plain SightSRI360 · on Wells Fargo85 / 100
EP284 Closest Alligator to the Canoe: How Transforming SOC Became P0 for Lloyds BankCloud Security Podcast by Google · on Agentic AI85 / 100Chris Black54 / 100Michelle Beyo60 / 100FintechTalks LIVE: Dave Birch70 / 100FintechTalks LIVE: Grace Keith Rodriguez63 / 100FintechTalks LIVE: Amanda Estiverne62 / 100