Reimagining Lending: Personalization, APIs, and the Future of Fintech with Matt Bivons, CEO of Canopy

In this episode of Banking on Information, Rutger van Faassen interviews Matt Bivons, Founder & CEO of Canopy, about revolutionizing lending through modular APIs, personalization, and modern technology. They discuss the future of lending, AI's role in fintech, and how businesses can prepare for a more integrated and data-driven financial ecosystem.

Key Words: Lending innovation, Fintech, Canopy, Modular APIs, Personalization in lending, Banking core systems, Post-origination software, 
AI in fintech, Data-driven financial services, Vertical software integration, Future of lending, Revolving credit, Loan management platforms, Financial institutions

Chapters
00:00 Introduction to Canopy and Its Mission
04:00 Understanding the Pain Points in Lending
09:28 Future of Lending: Trends and Predictions
13:13 Preparing for the Future of Lending

Transcript

Rutger van Faassen (00:01.398)
Hello and welcome to another episode of Banking on Information. Today, my guest is Matt Bivons who is the founder and CEO of Canopy. Welcome to the podcast, Matt

Matt Bivons (00:14.2)
Thank you so much, really happy to be here.

Rutger van Faassen (00:17.418)
Now we start every podcast with that same very important question. So here it is for you, Matt. WHY do you do what you do?

Matt Bivons (00:27.064)
So, I think there's two ways that we can answer this. Why did I found Canopy? and why does Canopy exist? And then why am I an entrepreneur? And they're slightly different, but definitely interrelated. So why did I start Canopy? I spent the last 10 years in banking and fintech and saw that borrowers wanted non-traditional loan products and yet...

Most banking products, most lending products are commoditized. The three-year loan that you get at Wells Fargo is the same that you would get at Bank of America. And even some of the fintechs that were part of my former life, trying to create a new user experience, still offered the three-year loan. And so I felt that the future of lending, which I know we'll touch on a little bit, wanted more innovative products.

People want personalization in the lending products they have. And that is true if you're a consumer or if you're a small business. which is primarily what Canopy does. And I think it's important just to give a little context on Canopy. We are an API first lending core and operating system for small business lenders. So if you are looking to launch and scale a lending program, you would use Canopy as your infrastructure.

post origination to help manage all the rules and policies that govern your product. So if you think about, you you using a credit card, but the back office software, the computation engine, statementing, all of the interest calculations, line item assignments all happen at the servicing and system of record layer. And so that's what canopy does. We've built a, a modern modular version of, what a lot of large incumbents have done, for the last 50 plus years. And so,

When I think about modern technology and I think about the actual borrower needs, there was just a massive gap in the market. And when I think about really any great product, great products facilitate better life experiences. And in lending, I think lending at its best helps facilitate those better life experiences. It helps you get through school with your student loans. It helps you choose the place that you're living with your mortgage.

Matt Bivons (02:50.582)
It helps you get to where you need to go with your auto loan. Lending it at its worst. It can obviously be very predatory and Max Levchin, the co-founder of PayPal and Affirm said this very well that most financial products are like trying to read a manual to a chainsaw that you've never used in before and have a high likelihood of cutting off a limb. So transparency in lending is really key. And we felt that

the transparency as it relates to real time information, the personalization to create products that are not just commoditized, again, going back to that three year loan didn't exist. And so we found this niche and that's why we do what we do to facilitate better lending experiences. The second part of it is why do I personally do what I do? And I think,

You know, the joke as an entrepreneur is you have to have a high tolerance for, for pain and rejection, which is definitely true. But, you know, I felt that building something from scratch and releasing it into the world and having other people use it and find some joy over it, is just such a rewarding thing. A friend of mine once told me he was a product manager at, one of the large FAANG companies.

Rutger van Faassen (03:50.549)
Mm-hmm.

Matt Bivons (04:14.818)
I won't name which one. But He said, you know, even on my best day, I could have my best day in an entire year. I'm still just a footnote in someone else's story. And so as an entrepreneur, not just as me as the CEO, but really everyone that works at Canopy, we ask them to think like an owner. And when you think like an owner, you are responsible for writing chapters of our story, not just a footnote. So,

making an impact is why I personally do what I do. The impact that we make to the world is Canopy.

Rutger van Faassen (04:48.714)
Yeah. Yeah. Yeah. No, that is, That is very interesting. And that's a, That's an interesting overlap. right? So not only do you want to make sure that the lending process is more personalized, that it is better, a better experience for the customer, but you also wanted to do it yourself. right? You want to be orchestrating that and designing it in a way that you felt was the right way to do it. So I think that is a great WHY. why you do what you do.

Now you talked a little bit about what you do. what is the number one pain point you solve for your customers? and maybe their customers?

Matt Bivons (05:28.504)
Sure. So, every single bank, known as an FI, every credit union, every community bank, even every fintech who is offering a lending or even a depository product needs a banking core. And traditionally, banking cores are run by very, very large, multi-billion dollar companies, FIS, TSYS Fiserv. And these companies are primarily built on a programming language called COBOL.

And that is a 50 plus years old. And there's nothing wrong with, with, with the programming language. Uh, But it is not meant for, for modern times. It is not meant to be integrated, uh, as an API first company. These companies are still on green screens. Imagine you're a 24 year old credit analyst and you just got a job at Capital One and you have to interface with a green screen,

to make changes to the credit line assignment. It's incredibly anxiety inducing because there are thousands of widgets that you have to try to navigate to and you don't know which one is going to impact things down the road. And this is ultimately why we go back to the first point of commoditization. It's because there's an inability to make change to these products. Now, that's not how consumers and borrowers live.

Think about if you're a small business owner, you own a pizza shop or restaurant. There's going to be seasonality to your restaurant. Maybe you came up with a new recipe and it gets written up with a great review and you get a ton of demand. You might need more working capital, right? Because your inventory, your staff, your rent, most small businesses take out multiple lines of credit. And if you're not able to fluctuate,

with the business needs, which traditionally you can't because it's so fragmented from a data standpoint and also so rigid from a database standpoint, you don't make those changes. And so the small business gets declined. so Canopy is an API first loan management platform. So after you decision someone, after you approve them, you create an account, in the Canopy system, and then all of the rules and policies that govern that lending product is within Canopy.

Matt Bivons (07:56.462)
Today we support 20 different types of lending products. They are traditionally classified in three categories, revolving cards, installment loans, and secured or asset backed products. Now that allows you to do a lot with those three groupings. You can create merchant cash advance, different lines of credit, securitizing against a credit card, against a HELOC or commercial real estate. You're able to create

AR financing, invoice factoring, lots of different financial products. And so we sell to banks and credit unions to fintechs as well as vertical software. So, one of our customers, Flexport, they're a global logistics company. They have a subsidiary called Flexport Capital, because if you're shipping goods from Asia into the U.S., there are many pieces along the way, and it typically takes 90 plus days to get from

the other side of the world. And if you are a supplier or merchant, you need working capital to continue to buy up inventory and run your business. So, Flexport capital provides that and they have a unique data advantage because they know exactly where the freight is along the way and can understand how they're going to get repaid. And so our software kind of software is accessed by the credit team at Flexport. It's accessed by the suppliers that use Flexport and then it's accessed by the merchants to make repayments within Canopy.

Rutger van Faassen (09:24.04)
Yeah. So the process after the loan or line has been approved, that's basically where a Canopy comes in and you provide a lot more flexibility than that typical three year loan that you were talking about. So for financial institutions that want something else than cookie cutter and something that is much more sort of personalized, that's where Canopy comes in.

Matt Bivons (09:46.83)
That's exactly right. So we call that post origination software. is where we come into play. So from creating an account all the way through charging off or closing your account, it's about 90 % of the life of the loan, 90 % of the engagement. That's where canopy comes in. Who uses canopy? Uh, The office of the C CFO uses canopy to close out their accounting books, uh, credit teams and risk teams use Canopy to, uh, simulate changes to their credit policy.

Rutger van Faassen (09:48.699)
Right. Yep.

Matt Bivons (10:15.032)
call center operators and agents use Canopy to answer questions for borrowers. Borrowers use Canopy to make repayments. And so some of the things that we do that differentiates us, obviously the flexibility of our data model, having multiple product types all under one roof gives a universal understanding of the customer. Most of these systems are very fragmented. They might even be on-prem servers for installment loans, another on-prem server for

revolving, you can't create hybrid products. We also offer an event driven system. So you're able to subscribe to web hooks. Every single action on our system is an event, it's an immutable database that has a full audit trail. So you can see all of these events and then we're able to publish those events, both to web hooks, to, to different data warehouses and data lakes for reporting. And then you're also able to build off additional workflows. And so,

one of the main differentiators of canopy outside of the flexibility and versatility of the product types we offer, it is in the servicing workflows and how we can connect to the other pieces of the lending ecosystem. Things like issuer processors, payment processors, accounting software, communication tools, all of these things are very complicated. Banks and fintechs already have a CRM. They don't want to migrate to a separate lending CRM. And so we connect.

right into Salesforce as an example, or right into NetSuite or Xero on the accounting piece. So as a platform, we're really an orchestration layer that sits on top of the other processes that you run your business on. We call them primitives. So primitives are really building blocks for lending. We'd like to say Legos for lending. And so as a system of record and banking core, as a system of record and lending core,

Rutger van Faassen (11:55.855)
Yeah.

Rutger van Faassen (12:01.636)
Mm-hmm. Yep.

Matt Bivons (12:09.036)
we are the heartbeat of a lending program, but there are a lot of other services that need to surround this that we integrate directly in.

Rutger van Faassen (12:16.592)
Yeah. Yeah. So that, that was exactly what came to mind for me, like an orchestrator, between all the different systems, between all the players involved in the servicing of a loan. even maybe self-service by, by the borrower. But you basically bring that all together. That sounds, that sounds very valuable. Now I'd like to do this thing called Futures Thinking. So we don't know what the future holds, but we can think about a possible future. So if we think 10 years out, 2035.

What do you think the lending market looks like? Paint me a picture of what your possible future could look like.

Matt Bivons (12:51.662)
Well, one thing I think is that we're still seeing develop is the fact that five, six years ago, that's when Canopy was started, there were a proliferation of consumer debit card companies. So Chime, Step, Current, Greenlight. the list goes on and on. And this was really a self-fulfilling prophecy because Issuers like Marqeta made it really easy to launch debit cards.

If you launch a debit card product, it's kind of like a square rectangle. It's very hard to retroactively move into a revolving or lending card, product, but lending products are where companies make, the most revenue. And so I do think that most of these companies, that, that are debit cards can't exist as a single product need to diversify into a multi-product company. We're seeing this.

with several of the international Neobanks, Nubank is already doing this, Monzo, et cetera, in Europe, launching into lending and secondary products. SoFi did this here in the US. That's very important so that you're able to create a high LTV and retention rate for your customers. So having a debit card only isn't enough. One of the challenges though is that until Canopy came along,

there really was a difficulty to create revolving cards and integrate with debit issuers. And so, we, we help a lot of the debit issuers become revolving issuers. So I think more debit cards moving into credit cards is number one, that's probably a short term future. I think that, you know, companies that are in payments today, vertical software payments companies like

Rutger van Faassen (14:44.42)
Yeah.

Matt Bivons (14:48.32)
Zelis in the healthcare space, Toast, obviously in the restaurant space. These companies are very well aligned to understanding the customer workflows and data in a particular industry, much better than a standard payment processor like Stripe or Adyen. would be able to do. And so, payments is a natural gateway into lending. And so I think we're just starting to see

Matt Brown of Matrix Partners has written a lot about this, where you look at vertical software and payments being the first entry point and then lending being the second wave. I think that is just starting to happen. Obviously some of the macro challenges of the last couple of years with rates being high and inflation, I think it made it a little bit more difficult from the cost to borrow against for some of these non-bank lenders to start lending programs. And as things start to stabilize,

I do think there will be more, more lending programs in the future. Lending is somewhat recession proof, right? Good times, bad times, like people are still lending. It just matters the rates that they lend against. But obviously, if you, if you are not a bank, if you don't have the deposit base, it makes it a little bit more expensive for you to borrow against. So, debit cards into credit cards, definitely a vertical software moving from payments into lending. then, Everyone is obviously talking about AI. and,

Rutger van Faassen (16:14.904)
Mm-hmm.

Matt Bivons (16:15.862)
You know, I think what is being missed in the AI conversation is most people are thinking about agents and the agent world as being able to have outbound or inbound conversations that would make call centers obsolete. That may be true, but today the software that you are interacting with

It does not know the additional data from a lending or system of record standpoint to be able to create actions. So pick any of the proliferation of AI, voice AI, agent tooling. Sure, I think they can maybe help with some collections. They might be able to say, hey, Matt, reminder your payment's due. on next week.

It's a slight step up from an IVR program. But if I were to call in and talk about a hardship or, want to restructure my, my loan or want to be able to split my payments, these systems can't do that because they don't understand the lending data. And so I think the adoption of AI in fintech at least beyond this initial use case, which is very basic, just the, it's like an advanced chat bot. You're, you're having a conversation with it.

I think it's going to take a couple of years to develop. Also the fact that in highly regulated industries like healthcare, like finance, you can't even have 1 % hallucinations. You need it to be a hundred percent accurate all the time. I like to joke that, you know, being in, in this industry is a little bit like coaching the, the USA men's basketball team, that, know, you're expected to win, but when you lose everybody complains. And so the system has to work. If it doesn't work.

Rutger van Faassen (18:03.297)
Right. Yeah.

Matt Bivons (18:06.51)
If you're getting bugs, if you are miscalculating people's statements or minimum payments, there are obviously significant regulatory risks there. so being able to get very specific and verticalized and take these, you know, large language models and customize them for a particular vertical, and then connect that to the actual underlying data, I think it will be the next wave of things and certainly will be.

the future of lending as I think, you know, several years ahead.

Rutger van Faassen (18:38.477)
Yeah. So blurring of the lines, lending and payments coming together, obviously impact of AI, but still human in the loop. What does that mean? So with that future, what do people need to do today to get ready for that possible future?

Matt Bivons (18:57.26)
Well, I think number one, know, working with a system like Canopy or other modern software that is flexible to allow you to build on is really important. The complexity of the banking world is because they have continued to duct tape and band-aid over very, very old systems. And then it's like trying to

you know, redirect an oil tanker. It's very, very difficult. You know, we're, still pretty small and nimble. And so we're able to, to build and execute at a very, very high, high velocity. And so not putting yourself in a corner, you know, again, that square rectangle example, if you build just a debit card, you then can't reverse into revolving. If you build just for installment, you can't, So trying to think about your, customers as they grow and, and, and

become multi-product from the get-go, I think is very, important. I also think that everybody should be working with some sort of co-pilot to help with productivity and understand AI tools as an augmentation to your day. That's very important. We do that at Canopy. Internally, we use it for prospecting. There's a bunch of additional efficiencies that we've gotten from that. And then...

thinking about who owns the data, like the data flow and Mark Benioff said this really well, that it's not the models, it isn't the servers or the infrastructure of Nvidia, it isn't even the UI, is who owns the data will be the ones powering the future of AI. And so thinking about the data flow of different businesses and where that resides is certainly a very, very big competitive moat.

that should be planned from starting now.

Rutger van Faassen (20:53.643)
Yeah. So make sure that you are more modular, more nimble so you can build for the future and think about your data model and how to structure that. I think that's probably a good point to wrap this up. Thank you very much, Matt, for being on the podcast and for thinking through the futures thinking with me.

Matt Bivons (21:13.718)
Absolutely, this was really fun. Cheers.

Rutger van Faassen (21:16.212)
Great. And until next time, choose to be curious.