Private credit plays a growing role in finance and this market segment now exceeds $2 trillion in value. As it has grown, so to have worries that it conceals significant risks to the global financial system. Unlike banks, private credit lenders operate mostly out of sight of regulators. Plus, companies that tap the private credit market tend to be at the riskier end of the spectrum.
A significant risk factor in the private credit market is the inherent difficulty in assigning a value to loans that don’t trade. Lenders often have little insight into their risk exposure because they don’t have a good way to monitor the performance of those loans on an ongoing basis. There’s not enough transparency. An additional concern is that the growth of private credit is attracting managers who are new to the asset class and lack experience with the risks and behaviors of this market.
While there are broadly applicable solutions for document ingestion already available, there’s a need for tools that help investors better identify the risks in private credit and track them over time. That’s where Boston-based EnFi comes in. EnFi (the name comes from Enhanced Neural Financial Intelligence) is leveraging AI to provide complex credit analysis and risk monitoring services for lenders. It enables process automation and improved risk analysis during underwriting as well as ongoing monitoring of performance, including the monitoring of covenants. EnFi is designed for any complex credit — capital call lines of credit loans, mid-market, mezzanine, venture debt, agricultural, yacht, and more.
EnFi was founded in 2023 by Joshua Summers and Scott Weller, both of whom are serial entrepreneurs. In June of ‘24, they raised a $7.5 million seed round led by Unusual Ventures with participation by Boston Seed Capital, Argon Ventures, and Impellent Ventures. EnFi launched in private beta several weeks ago.
Q. Joshua, how did the Silicon Valley Bank fiasco lead to the founding of EnFi?
A. I’m a serial entrepreneur. My cofounder Scott is also a serial entrepreneur. We’ve spent our careers solving big problems around big data. I spent the last twenty years in advertising technology; Scott in customer data platforms, loyalty, gaming, and all kinds of other things. We’ve been very fortunate — we’ve had some good successes and these successes led to us becoming angel investors.
After selling my last company, I started an angel group with a half dozen really great entrepreneurs called TBD Angels. TBD angels is now about four and half years old and we’ve invested in around 80 companies. We have about 300 members now. It’s entrepreneurs investing in entrepreneurs. We do what founders do, which is to give back to the startup ecosystem because it’s been good to us. Scott and I have been co-investing through TBD Angels for some time and really got to know each other.
When Silicon Valley Bank and First Republic had their fiasco, it was a frenetic 72 hours for both of us — and really for anyone in the startup investment world: for VCs, for angels, for everyone. We had dozens of founders calling us panicked. When a bank run happens, you don’t want to be the first company to participate, but you really don’t want to be the last.
We got these calls from people asking for help. It was let’s connect you to this person at this bank; let’s connect you to this person at that bank. Let’s see if we can open accounts in the fastest way ever because usually it takes time to open these things. Over the course of the weekend and into Monday, we moved a ton of accounts for lots of different companies leveraging our connections and experience.
There were a bunch of companies where we ran into roadblocks. These roadblocks existed for companies that held debt at First Republic or Silicon Valley Bank, because of covenants that restricted where they could hold their cash. Those covenants make sense for lenders. Lenders put them in place because they are a great way of assuring they have a primary source of repayment. It’s a risk mitigation strategy and there’s nothing wrong with that. But in this situation, those covenants were a hindrance to solving the problem.
Scott and I stepped back and asked ourselves what we could learn from this experience. We started calling friends in banking and in private lending and asking questions. We set up literally four dozen conversations across small to large lenders, private and institutional, just to understand the nature of covenants, the nature of these requirements, and the nature and process of lending. We found a lot of consistency across all types of lenders. Our research revealed significant inefficiencies in financial workflows and risk management. These inefficiencies led to prolonged processes to underwrite and prolonged processes to originate. Most institutions had a limited view into the risks that exist after a loan is issued due to human-powered analysis and monitoring and a lack of resources to go deeper. You can’t re-underwrite a loan every single day because it’s humans who are performing those actions.
That led to our thesis which led to us creating EnFi. The best part of that process was before we even knew we were creating a company we were able to talk to dozens of experts and do this discovery process with no agenda. With every company I’ve started over the years, I wish I had done this. It’s amazing what you can learn from experts who have been in the industry who know where the bottlenecks are and where the challenges and the blind spots are and because you aren’t trying to sell them anything they’re willing to share. It was a great process.
Q. How do you define your target market?
A. Today, it’s primarily institutional bank lenders, private credit lenders, and specialty lenders. We concentrate on complex credit scenarios and exclude more prescriptive lending types like mortgages and auto loans.
What we want to do is look at bespoke lending concepts and figure out how to build technology solutions — accelerants — around those. We’re not looking at the largest lenders but at those which are small to medium sized. They don’t have big in-house teams and they’re looking to increase their ability to understand risk.
It’s U.S.-only to start. That’s where our connections are, that’s where the market is raring to go. We’re a small team today but we see this problem everywhere. It’s not a U.S.-only problem.
Q. How big is your market?
A. Always a fun question but a hard question to answer. There are trillions of dollars in the institutional and private credit markets. There are tens of thousands of companies issuing loans of various sorts into these markets. The opportunity is huge and the momentum is there.
Private credit is on fire. Regulators are starting to get very involved; they’re trying to understand this side of the market but they’re still somewhat blind. In addition, AI has opened this new set of discussions. In two years, it’s gone from, “No way are we bringing that into our lending approach,” to now, where 60%, 70%, 80% of the institutions we talk to say, “Absolutely. We know it’s a must. We just have to figure out the appropriate way to do it.”
Q. How do you plan to reach your prospects?
A. One of the things I’ve learned is that you always start with founder-led sales. The best reason to do that, is that as a founder you learn so much from these conversations. You learn where the market is, where your product has gaps, who the competition is, etc.
So, we’re starting with founder-led sales. It’s really through basic networking and conversations like this. Jim, I can guarantee that you’re going to write something up and I will get one to three to five inbounds and some of them will look great. It’s happened with everything we’ve put out. Someone will say, “I read this piece on you and it’s exactly what we need.” This is what we do to start, but that’s not how we scale sales. It takes you through your first ten or twenty customers.
To date, we are focused on finding customers who can help us prove out product-market fit. Once we get confident about that, we will expand our muscles around sales and go-to-market, we will hire the right team, we will think about the right ways to market ourselves, we will spend time at the right conferences — we’ll do all the standard things to go from being an in-the-weeds startup to being a company that can scale and grow. I would expect that to be a next year activity for us.
Q. Is there evidence that lenders are dissatisfied enough with their current risk monitoring solutions that they’ll be willing to alter their current workflow and integrate a new product?
A. Absolutely! There is so much evidence! I look at this as two questions. First, what is the evidence, and second, what are the barriers to integration?
On the evidence side, we found that many lenders really don’t have what would be defined as risk monitoring solutions. They have workflow solutions that allow them to use standard methods for thinking about risk, put covenants in place, and then monitor those covenants. That’s all well and good, but that’s not really a risk solution. It’s a bunch of Band-Aids. It doesn’t allow them to truly understand and assess the risks throughout the loan lifecycle.
These lenders understand they need to go to the next level, and they are embracing these conversations. We are talking to chief credit offers, chief risk officers, CEOs of small lending organizations. They are honestly very hungry to talk to us because they recognize this challenge, and they recognize that more regulation is coming to the space. We are not running into any barriers to consideration.
A lot of people historically believe that when you sell into institutional and private lenders, there is a long, drawn-out integration phase. They’re thinking about traditional loan operating systems or core banking systems. What we’ve done is rethink the way this software can be sold into these institutions as value-add on top of those other systems. We’re not replacing your other systems. We bring value on top of them, and we do that through something call a “minimum viable integration”. Everyone talks about MVPs, we believe in MVIs.
How do you sign a deal with us and within hours or days begin to use the product and realize value? We’ve designed the platform to allow for that. We don’t need a multi-month onboarding process. That doesn’t mean that there aren’t things we can do to make it even more valuable. Of course there are integrations into the core banking system, of course there are integrations into the loan operating system, but they’re not required to begin using our platform. You can use it Day 1. Then once you do that you can unlock the additional power of the platform. We’re really excited about this. It’s an actual rethinking of the way software can be sold into these financial institutions.
Q. Are you able to demonstrate improved risk monitoring at this early date?
A. That’s a tough question. What we can demonstrate is efficiency gains. Over time we should see 80% – 90% improvements or more for the three key stages of loan operations: origination, underwriting, and portfolio management.
Our immediate KPIs for customers are aligned with efficiencies in their workflow (how long to gather the information to underwrite or monitor the loan, how quickly to assess compliance with covenants and other obligations, how long after borrowers submit data does it take to understand where risk is moving, and how long does it take to produce risk analysis and credit memos, reports, etc.).
What we can’t demonstrate today, because it takes time and it takes data, are reduced loss rates. I’d love to have you ask me that part of the question next year, because I think we’ll see improvements in loss rates by using AI to open the aperture so you can see more risks, and process more risks, and understand more risks. We just don’t know what the improvement rate will be. We need more time with our beta customers to allow loans to matriculate through the entire lifecycle to measure this accurately.
Q. How do you ensure compliance with client lending policies?
A. We’re not making the lending decision. EnFi is a platform that creates assistants to accelerate the effort to make lending decisions and to monitor those decisions. What we believe is that we give you the power of 10,000 analysts sitting by your side when you are working on a loan.
We extract information from credit memos and loan agreements and other types of legal documents, and we infer the obligations and requirements seen in those. We can also understand the decisions that went into making that loan and how the risk was understood. All of that, AI can pull out, summarize, and use.
The first part of our product, which is in the market now, is focused on portfolio management (the third phase of the loan lifecycle). It’s about monitoring loans once they’ve gone live. We call it continuous risk analysis. Effectively, it is bringing the power of re-underwriting the loan every single day throughout the life of the loan. That’s the holy grail — always have the freshest understanding of all the risks that exist for your loans.
We are quickly moving into origination and underwriting. The cool thing there is, we can ingest your credit policies. In ingesting those, now we can start to tell you when there are anomalies. We can surface that as “here’s something you need to pay attention to,” or we can surface that as “here are ways you can mitigate this”.
Q. You say you provide “real time” monitoring of loans. How real time is real time?
A. You’re right to ask this question. We call it Continual Risk Analysis. Essentially there are four key types of data that are part of that analysis. You have borrower supplied information which is usually reported monthly or quarterly over the life of the loan, you have public data, which you can think of as government filings, market level information, etc., you have private data — think information collected by third parties around companies, people and other aspect of the type of loan being processed — and you have lender supplied data (think credit policies, reports, portfolio performance, etc.).
Real-time in the context of EnFi means we ingest it as it is received and within minutes the system is updated with the latest and greatest view of the loan and the monitors that exist around it. Alerts are fired off if necessary to inform as changes happen. Real time means as fast as the data is changing.
Q. You’re building end-to-end ‘agentic’ workflows. What does that mean?
A. We hemmed and hawed about putting this on our site. AI agents are autonomous software entities designed to perform specific tasks or achieve particular goals within a given environment.
In the context of EnFi’s system, AI agents are specialized components of the platform that work together to handle various aspects of the lending process and very nuanced pieces of that lending process. Agents can perform specialized functions like planning, or compliance analysis, risk analysis, data gathering, summarization, etc.
Individually and autonomously, they perform these tasks, but they also work together and communicate with each other. An agentic architecture means building a bunch of agents and having them interact, in our case, through something called swarm intelligence, where their combined efforts produce sophisticated outputs and insights for lenders. We use these agents to execute workflows in the lending lifecycle.
It’s a really cool concept. Swarm intelligence harkens back to the way bees operate. Each bee in the nest has its own job, but if one bee doesn’t survive, another bee picks up that task. The bees all work together to create this unbelievable environment. That’s what an agentic architectures with swarm intelligence does. That’s how we built the platform to get more done with AI.
Q. It’s one thing to monitor the performance of a loan, and another to monitor the performance of a loan portfolio when each loan entails intricate credit scenarios involving multiple risk variables. How are you approaching risks at the portfolio level?
A. Portfolio management is a key use case for our solution, addressing both workflow and risk analysis challenges, and it’s the first use case we’ve tackled.
On the workflow side, we automate the coordination and analysis of numerous documents from borrowers and other sources, typically received monthly or quarterly. Our AI automatically reads these documents, analyzes them. It does things like spreading financials and assesses their impact on obligations and covenants for each loan and across the portfolio. This allows us to surface outliers, enabling portfolio managers to focus their attention where it’s most needed.
This is a massive efficiency gain. Think of it like this: every month or quarter, you have a portfolio of loans. All those loans dump all their data within a 24- or 48-hour period. You have now hundreds if not thousands of documents to get through. That takes a huge amount of human capital to perform, but we do that in minutes and can say here are the 12 things you need to look at.
Currently, we’re focused on enhancing our risk monitoring capabilities. We’re moving beyond basic compliance checks to incorporate external factors like interest rates and market-level data, along with borrower-submitted information such as broader financials. Not just matching the covenants to their financials but looking deeper to understand what’s truly going on. There may not be a covenant around days outstanding on receivables, for example, but if we notice accounts receivable move from say 65 days to 90 days, that is a risk factor we want to be able to understand. Our system analyzes how these factors can lead to changes in risk ratings, applying this analysis across each loan in a portfolio to produce necessary portfolio-wide views. This provides a more dynamic and comprehensive risk assessment.
Looking ahead, we plan to further integrate our solution with lenders’ specific credit policies and portfolio risk strategies. Lenders have different risk policies for different types of loans, and if you think of the portfolio as applying those policies as percentage allocations, we can do some really interesting things.
Our goal is to develop a system that can assess how a portfolio moves within these guidelines and automatically surface relevant insights. This will provide a truly holistic view of portfolio risk, accounting for the complexity of each individual loan while aligning with the lender’s overall risk management approach. We’re also exploring predictive analytics to help lenders anticipate potential issues before they arise.
Q. Are all lenders comfortable with a SaaS solution, or are you finding some still want things on premises?
A. It’s important to note that we are focused on small to midsized lenders. They don’t always have the infrastructure needed to do everything in house. It’s expensive to do it and hard to maintain it. They have really embraced these hosted platforms.
It’s important to also note that we have built a flexible solution that can be hosted in the cloud in our environment or hosted in their private cloud environment. That’s where things are today. It’s a private cloud, it’s not an in house managed platform. We can work in either of those environments.
Regardless of the deployment model, we adhere to stringent security protocols to ensure the protection and integrity of our clients’ sensitive information, aiming to provide a solution that enhances lending operations while aligning with each organization’s unique security needs.
Q. You just finished raising a fairly large seed round at a time when funding for fintechs is down. What was the fundraising environment like? How quickly did the round come together?
A. We were very fortunate to be able to gather a great set of investors and a strong lead to make this round happen pretty quickly.
Scott and I have built multiple companies. We’ve raised lots of capital previously, so we have good relationships with lots of investors. We were very lucky to find a great set of investors who had theses in this space who believe in us as entrepreneurs. The momentum was there. If you’re bringing the right AI solutions into the market today, with the right story, the capital is there.
Before we were even a company, we started to socialize the idea and get feedback from VCs we believed would be a good fit for this — not raising capital but trying to understand the dynamics of funding an idea in this space. Through that process we found a small number of investors that were a good match for this type of company — deep expertise in financial services and AI, founders on the investment teams. This is a pet peeve of mine. Investors with a financial background are great. They bring a good perspective. But a founder who has “been there done that” brings a very different type of perspective. A history of supporting companies is important, not just when things are going well but helping them when times are tough, because every startup is a rollercoaster.
The investors we partnered with were key to building out our thesis. They helped us before the funding discussions. They delivered value before the term sheet by connecting us to potential lenders and helping us refine our ideas around the problem and the go to market.
The timing was good for an investment in this space. We oversubscribed our initial fundraising goal by a considerable margin and had to turn away some incredible investors. Eventually you get to a point where the economics don’t make sense. You can only take so much money relative to the valuation. The investors we ultimately partnered with included Unusual Ventures (Lars Albright), Boston Seed (Nicole Stata), Argon Ventures (Bob Mason), and Impellent Ventures (Phil Beauregard).
Q. Was it a SAFE or was it a priced round?
A. A priced seed round.
Q. Was there a difference in how the deal was viewed by investors in Silicon Valley versus those in Massachusetts?
A. Yes and no. Unusual Ventures led our round. They are technically a West Coast VC but they have a Boston office and our lead works out of this office. As a company that has created a bi-coastal strategy for investment, they’ve thought through some of the nuances of operating in both areas.
The reality is that, 10 years ago, the differences between West Coast and East Coast were far more pronounced than today. East Coast has started to think more about how they compete with West Coast money. There were some nuances, places we saw differences, but these are all seasoned investors. They’ve done this before, conversations were able to happen, and we found a happy medium.
For us, if we had taken $1 million or $2 million in a pre-seed — and there were plenty of offers for that out there — we probably would not be able to achieve the KPIs that will allow us to get to the next level and the next raise. This is not an easy problem to solve. It’s going to require some capital and some time to get it right. We knew that, we presented that story, and the investors we brought in got it, understood it, and wanted to participate.
Q. You recently wrote about the lessons you learned at your first three startups that you’re now applying at EnFi. One that I’d like to ask about is the need to raise more capital than you think you’ll need. Having run out of capital at my first startup, this is advice I always offer when talking to founders. A lot of founders fixate on dilution and don’t judge the risk of running out of capital correctly.
A. You, with a founder background, understand founder math. We all do it: what’s the dilution relative to the capital in, what’s this mean five, 10, 15 years down the road? It’s an important lesson I and so many other founders have learned. It’s not always up to the founder. You can’t always say you want to raise more money and find the money is there. That’s a very privileged position to be in. We were fortunate to have more demand to participate in this round than we expected and that allowed us to over-index on raising more capital. Many times, the opposite is true where you have to do more with less.
In either case, diligent use of capital is critical – what you do with the capital you’ve raised and how you make your dollars stretch. Use them if they accelerate your delivery of KPIs but don’t assume the next round is always going to go the same way as the current one. Even if it all goes perfectly, and you massively outperform — you did everything you said you were going to do and more — the money is not always there. Markets change, VCs change, competition changes. A lot of it is outside of your control.
Taking more money comes with more dilution. If you can do it, I recommend the tradeoff. The reality is that many years from now when you have finally achieved an exit, the difference of raising a little more at a little more dilution in the early days versus a little less where you may ultimately have to raise more in the future to make up for it, is probably a wash. I always say err on the side of more capital.
If the exit is a good one you will do well in either case. If it is not a good one, the difference there is moot anyway.
Q. What have you seen from the private beta thus far?
A. The early indications are exciting but it’s still early. We know that we can dramatically accelerate the time required to monitor a portfolio of loans. Efficiency is what’s easy to see. We also know we have a lot of work ahead of us to bring more understanding and deeper analysis to the identification and tracking of risks at the loan level and at the portfolio level. This will be a great question to check in on at the end of the year.
Q. What will your offering look like a year from now?
A. We will have solutions in market that work across all three major phases of lending (origination, underwriting and portfolio management). We will have enough time with beta customers and will have run enough loans through the platform to confidently state what the system is capable of. We will know where we need to double down to continue to build out the best risk platform for complex credit and the most complete picture of that risk over time with Continual Risk Analysis. Anything more than this will be hard to predict as this space is evolving so quickly that whatever we predict today will be out of date by next week.
Q. You are headquartered in Boston as am I. What is the Boston fintech ecosystem like in your experience? What are its strengths and weaknesses? What do we need to get to the next level?
A. Boston has some really incredible things working in its favor. First, we have talent generation. We have MIT and Northeastern and other great schools producing talent that are going deeper into AI and other really interesting technology concepts that are directly applicable to fintech.
Then we have some really great VCs and Angels in the Boston area. This is critical to financing that talent to go build new and innovative solutions in the fintech space. In my opinion, we are a Top 3 market for that. Silicon Valley, New York, Boston — we deserve a place in that discussion. We’re not the top, but we deserve a place. Because of the match between talent and funding, we have a lot of fintech companies coming out of the Boston market.
Finally, we have some really incredible companies in Boston supporting the ecosystem that are either headquartered here or have a major presence. Companies like Fidelity and other large institutions as well as break-out startups like Circle, Toast and other companies that have achieved exit velocity. All of those generate serial entrepreneurs. Almost everyone I worked with at my prior company, which was acquired by PayPal, went off and started another company.
Beyond that there are hundreds of startups and lots of ways for them to interact, collaborate, and learn off one another.
In addition, our state is thinking about how it supports the innovation economy through different outlets like MassTech, Massachusetts Fintech Hub and MassChallenge. They’re really important for continuing to nurture the fintech startup ecosystem.
It is not perfect. We’re not number one and we’ve got work to do. We can do more to continue to develop Massachusetts into a stronger player in fintech. We are not the financial center like NY. Everything we can do to convince large financial institutions to put more people here in Boston would be good for generating fintech startups in Boston. We don’t have as much access to capital as Silicon Valley. We need to close these gaps by doubling down on ways to support the ecosystem through policy, engagement and investment from the state. There are a lot of great things but we’re not there yet.
Q. Are you hiring?
A. Yes. We are specifically looking for people who have deep AI engineering experience. We’re looking for senior AI talent. Not a researcher but a practical engineer who has been doing AI for years. We have a pipeline of great candidates, but we would love to see more.
We’ll soon be looking for someone to come on to help us with our go to market efforts, specifically sales into private lenders. Someone who’s done $0 to $5 million who understands what it’s like to get out there and have the early conversations, land the first customers. So someone who can take it from founder-led sales and build a sales organization.
If you are passionate about this space and interested in what we are doing, we would love to hear from you. Reach out over LinkedIn.
Q. What else would you like us to know about EnFi?
A. One thing we didn’t cover completely is what makes us EnFi. I’ve spent my whole career thinking about how to build good companies. That doesn’t mean just outputting a great product and keeping your customers happy, scaling sales and finding an exit. “Good companies” means people want to work with you. It means you’ve figured out how to provide the necessary support mechanisms, mentorship, growth opportunities, work-life balance. That’s super important to Scott and to me. Incredibly important. We want to be held accountable to our employees as much as we want to be held accountable to our customers and as much as we need to be held accountable by our investors.
Yes, our number one thing is how do we generate a great return for our investors. But what we think about is how we build a lasting company where people want to work. On thing we’ve noticed, because we’ve been focused on that very thing for decades – both of us – is that employees we’ve worked with in the past jump at the opportunity to work with us again. That means more to us than you would ever believe. It tells us that we’ve done something right and that’s what we’re trying to do here. We want to build a great company and be the company everyone in Boston wants to work for because were doing cool stuff, but also because we care about them as employees, and we’ve built a great place to work.
# # #
