How Lenders Actually Assess Small Business Risk

What We Need to Grow, Episode 026: Risk Is a Movie. A Founder Talk conversation with Juana Attieh, Co-founder and Chief Strategy Officer of AMLOK. Cleaned for readability; the words are the speakers' own.

Caleb Pedosiuk: Juana, first off, thank you. Some of the most important problems I'm seeing right now that teams are looking to solve are in the financial tech space. A lot of automation has come alongside promising a lot of things. You see people automating outreach, you see admin tasks. But when it comes to financial problems to be solved, this is a really important problem, because if you get it right you have the ability to essentially be adding profit to the bottom line, and if you get it wrong there's a lot at stake. So I know that businesses are looking at this, the ones that are paying attention, and are really eager to find partners that are solving this well. If you can share a little bit about AMLOK, you're the Chief Strategy Officer there, share a little bit of what you guys have built, and who it helps, and why it is that this actually matters.

Juana Attieh: Of course. Thank you very much for having me, I'm very excited to be here. You're very right. Data is becoming a lot more accessible, but that doesn't mean that there are many ways for us to be able to interpret it. This is across the board, but more specifically when it comes to financial data.

So what AMLOK is, is we're building risk intelligence infrastructure for financial institutions. The simple way I'd like to describe it is, lenders right now are still making capital decisions from fragmented snapshots of a business, while the business itself is constantly changing. They're still counting on snapshots and very stagnant data. So we help turn that fragmented financial data into a much clearer, continuously updated view of the business and its risk.

The problem, really, is that a lender right now might ask the business for a bank statement, or credit information, or accounting data. A bunch of PDFs that still don't necessarily tell them who the business is actually transacting with, what their obligations are, if there's hidden debt, if there's concentrated revenue. This risk profile is ever changing, but right now there's no comprehensive way for the lenders to see this about the businesses they're deploying capital to.

So AMLOK acts as this intelligence layer, where we solve what we refer to as four Vs.

The first is the visibility problem. Can a lender actually see what's going on inside the business at all times? Financial data right now is really opaque in its nature, and specifically in Canada it's very interesting. For example, we have over 30 percent of businesses now transacting through Interac e-Transfer, which is a peer to peer transaction. However, the way the data descriptors are when the transaction is posted is very opaque. The lenders cannot understand where this transaction came from. It usually is either a name, or an email, or some sort of truncated descriptor that we have no idea where it comes from.

Payment processors, too. A lot of businesses depend on online payments as well, and the way these transactions are posted is a lump sum, just one deposit. There are a lot of underlying signals that aren't entirely captured when the business goes to seek capital from a lender right now. So we solve this by connecting to the entire business's financial ecosystem, whether it's their banks, their payment processors, their e-commerce, their accounting software. We resolve and enrich this data to give the lender a more comprehensive view of the risk profile of the business they're going to deploy capital to.

There are a lot more nuances there. It's interesting for Canada, for example, we have bilingual issues where you have French names and English names, and we have no centralized hub where all the businesses are registered. So there's a lot of fragmentation of data, where when a lender goes to deploy capital to a business right now, they only get a really small slit of the entire picture that they could essentially get.

And people could say that this is being solved right now with open banking, which is really massive for Canada. Open banking does solve the access to data. However, it doesn't solve meaning, interpretation. So we help on the visibility aspect of that.

The second V is veracity. Can you really trust this information that you're looking at? On one end, AI has helped us in many ways, but it also equivalently made it a lot easier for you to fake documents. This is as old as time, where as soon as banks started giving out money and requesting documents, people were faking documents, or finding different ways to go around the system. So veracity is, how can you not just rely on a single source of documents as your truth, but be able to cross reference across all these different data sources that you're able to get, and also use the new advancements in technology that can backtrack where information comes from and how it's been edited. So we really want to solve the verification aspect, the veracity problem, that's getting augmented now with AI.

The third issue is velocity. Okay, we have all of this data, we've solved the idea that you can trust the data we're giving to you. But right now lenders have a few people that manually do the underwriting and the matching. So imagine we're giving them all of this new data that they have to digest. It's going to be a lot slower than it is already.

So what we do at AMLOK is we automate a lot of the workflows around the decision. The main idea is that we work with financial institutions, and our goal is not to replace their systems or replace their workflows, but to embed within them and augment them. So we embed inside the entire life cycle of a lending workflow. We help them automate everything around the decision. Whether that's how you structure the information, how you configure rules, how you map it to underwriting to trigger the next steps, who it goes to. So we can help at least with those cumbersome workflows that make the process a lot longer, while still keeping the decision itself with the credit officers, wherever they need to be.

One of the biggest things with AI right now is the explainability aspect. AI is a black box, essentially. But now we're seeing a lot more regulations come out where, if you are using AI to help with your decision making, there needs to be an explainable aspect to it. We take this very seriously at AMLOK, where we just act as this explainable intelligence layer, but the decisions and the policies remain with the right parties.

And then the last V is vigilance. I especially like this one, because we don't see it existing as much now, but I think it's the missing piece in how we can bring everything together. It's, okay, what happens after the money is deployed? Right now the lender business relationship kind of ends there, and then there are periodic reviews of, let me just see how the business is doing. Or I'm not going to know if anything's happening with the business until they don't make a payment, or they default.

So what we do for the lenders is allow them to continuously monitor the health of the business. Like I said, the business is an evolving entity. You can get the money now, you can lose your biggest customer, you can go and get another loan. There are all of these risk signals that could affect the risk profile of the business even after you give it money. So we help the lenders see continuously the full picture of the financial health of the businesses that they support, so that they can be proactive rather than reactive. They can predict defaults, they can detect stacking, they can maybe make dynamic structures that really benefit the business as well.

Right now a lot of terms are made with the assumption that the business is going to default. But with this, since you have a full picture and a continuous understanding, you can give the business better terms, better interest payments, however it makes sense for the business and the lender, and then you can have repeat customers, you can retain the businesses that you work with.

So the way we like to tagline it is, risk is essentially a movie, but right now it's treated as a photograph. We want to keep it as this continuously evolving movie. But this is it in a nutshell.

Caleb: So who would you say is your primary target client? Are they investors that are looking at companies to improve them, make them more profitable? Are they banks that are being approached by these SMBs looking for financing, and they're mitigating their risk while wanting to turn a profit on a loan? Or are they the actual businesses themselves that are looking to become more attractive to investment or to banks?

Juana: Right now we really are focused on three types of financial institutions. Credit unions, because they're member based financial institutions that have a lot to compete with right now, so we're really looking to augment their processes. Alternative lenders, the revenue based financing, the get approved in 24 hours type of institutions. And then banks, of course, to help them with their lending workflows and their underwriting and approvals.

But through working with those institutions, we also offer white label dashboards to their businesses, where the business can continuously monitor their financial health. On top of that, we've essentially mapped all the different grants and tax credits and opportunities that the businesses could be eligible for. This helps the business get into better financial standing, which then helps them be in a better place to get loans.

Caleb: You mentioned a movie, which I love, the comparison between a photograph and a movie. If you were to use the story analogy and play out the logical flow of how a customer hears about you and comes into your world and is integrated, onboarded, and has a functioning version of your dashboard or tools, what would that look like?

Juana: I can give you what we're working on with a couple of design partners right now. Essentially they pick out one lending product that they use, whether it's a line of credit or a specific type of revenue based loan, and then they have three options to work with AMLOK. Either they use our dashboard, which gives them comprehensive use of everything, or they can use our APIs, so they can integrate everything that we offer right into their workflows.

So the way it happens is we work on one workflow. They can also create how they want the businesses to be onboarded. The businesses that come to them will then connect their entire financial ecosystem, whether it's their Stripe, their Shopify, whatever other platforms they use. And then we essentially have a graph, a proprietary LLM that we've built, that maps out all of these different entity relationships and assesses the business's risk.

And then the partner or the customer, depending on their workflows, would be able to use this data to make different triggers. Either they configure it to certain underwriting policies, or if this happens then it would be sent to this department. So we make it really seamless and simple for the institution to be able to assess and digest all this fragmented data using our graph, use the outputs within their workflows, make decisions accordingly, and then monitor everything, whether within the systems they already use or using our dashboard.

And if they want to take it a step further, for credit unions it's specifically interesting, because credit unions have a different type of relationship with their members. It's more intimate and it's more long term. So they can offer these businesses a dashboard where they can plug in and continuously monitor their health, or get grants, or be in good financial standing, be reminded of certain payments. So we embed into the entire lending life cycle.

Caleb: What has your take been on the adoption of some of these tools? You mentioned design partners, which are so important at this stage of ensuring product market fit. Do you find most of these institutions are eager to find someone to work together with on solving this problem? Or are they more reluctant, perhaps strictly wanting to use an API, or just having a default position that pushes back against having a third party partner?

Juana: We've seen it all, to be honest. Obviously with new technology now there's a lot of, it's like a race. You really want to be in, but at the same time you don't understand the risks that are associated with it. And there is a lot of risk with using a third party. This is why we're really vigilant throughout what we do on the explainable aspect of it, and the research documentation, and our documentation in general.

I will say that there is a lot of eagerness, because a lot of these alternative lenders and credit unions do understand the need for adopting next generation solutions in order to stay competitive and remain relevant. But with these institutions it's obviously a longer sales cycle. There's a lot more compliance, there are a lot more considerations that you have to take into account. I would say the eagerness and the need is definitely there, mostly also because open banking is just around the corner, hopefully. And with that comes a lot of other issues that we might not have thought about. So we're really positioned to support all of these potential solutions that we can offer.

Caleb: I'm not sure if this is a fair question, but I'm curious. If you had to pick between having everyone in one category adopt your product, so you could snap your fingers and have a significant portion of VCs, those with money that are looking to invest, use your product as the gold standard and then require the deals they're working with or their portfolio companies to all integrate. Or if you were to have companies that are looking to use advanced tools to leverage things like grants and savings advantages. If you could pick between the two, which would you pick?

Juana: I have two answers to this question. I have my emotional perspective, and then I have the more founder hat, logical, scalable part of it.

I'll start with the emotional one. I'm very attached to supporting businesses. We've been in SMB financing for three years. SMBs are the backbone of every economy, as people say, but they are the least supported. So I definitely have an affinity towards supporting businesses as much as possible. But B2C is always really difficult, so B2B is really the smarter approach, because then it becomes B2B to C. That's where the split comes from. But our hope is that by going through these institutions that have portfolio companies, we are helping everyone across the board.

Caleb: Last question. What would you say, if you had to put your finger on one specific lever, is your greatest need right now for company growth?

Juana: We've been bootstrapped for three years, and we just kicked off our fundraising round. It's a small pre-seed round, and really the goal is just to go to market faster, turn those design partners into actual customers, acquire more customers. So I would say that's really our biggest need right now.

But more than that, I think it's really just continuing at the pace that we're at with the same conversations, and being introduced to other types of institutions, potentially other pathways that we might not have thought about, to mitigate that whole long sales cycle problem that we're discussing. But I would say right now it's really our first round of investment, that could just get us where we want to go faster.

Caleb: And the best place for others to find out what you're working on, what would you say?

Juana: I would say LinkedIn is our most active right now. My personal, and then we have our other founding team members, and our company profile as well.

Caleb: Awesome. Thank you so much for sharing, and I look forward to seeing how you guys grow.

Juana: Thank you so much for having me. I appreciate it.