The automotive industry has long relied on FICO^® Auto Scores as a method of quantifying and evaluating a borrower’s creditworthiness.

It’s the industry standard for determining whether or not an individual will repay their auto loans, and it allows lenders to make better business decisions. Unfortunately, even with the iterations of this standard over the years, Aaron, James, and Paul observed that there were severe process inefficiencies and a lack of data-driven decision-making when it came to the overall auto loan credit process. And they wanted to do something about it.

With extensive experiences in the financial services industry combined with data modeling, the three came together in 2020 with the idea of using AI in order to provide loans to the subprime market.

Subprime borrowers often face difficulties obtaining loans due to a higher risk profile, but with an in-depth knowledge of this sector, more borrowers could be included in the loan process and lenders could operate more profitably.

With a machine learning model that could train on tens of millions of outcomes, lenders could assess a borrower’s creditworthiness better. The FICO^® Auto Score would become only one piece of the puzzle. With AI assessing larger pools of relevant data in real-time, subprime borrowers, who previously would have been boxed out of the loan process, would now have a chance to get one, and lenders would be able to perform better risk analyses and make data-driven decisions.

Aaron, James, and Paul founded Karus and built out the suite of AI tools to analyze the necessary information for the underwriting process. By undertaking a proof of concept using public portfolios, Karus aimed to demonstrate the effectiveness of their risk analytics platform. They realized that existing platforms were outdated with interfaces that were difficult to understand and would not work properly for their needs.

So, at this point, they needed to build a platform that would transform Karus’ AI algorithms into a user-friendly, front-end interface that visualized the complex data in a way that lenders and financial institutions could interpret, analyze, and make informed decisions as quickly as possible.

With their proprietary AI model in hand, they approached OAK’S LAB to build a platform from scratch that could one day transform the subprime auto financing market.

To build a state-of-the-art platform that would convert complex, algorithmic data gathered by Karus’ AI model into sophisticated visualizations that are both functional and easy to understand so that lenders, dealerships, and institutions could monitor the real-time performance of loans and analyze price portfolios.

We needed to develop a solution that would improve the underwriting process and ultimately increase the accessibility of subprime auto loans. We faced the challenge of taking all the data from Karus’ AI model and making it presentable and actionable on the front end for our users, regardless of their technical knowledge.

In order to do this, we needed to ensure that the platform would be easy to use so that dealerships and institutions could navigate it quickly in order to make faster loan and portfolio decisions. Alongside a number of complex challenges that would come with this project, we needed to design a compelling UI that aligns with the brand of a leading-edge auto finance startup. Additionally, the UX had to allow users to see real-time data updates, and we wanted the entire platform to be web-based for greater accessibility compared to a standalone computer file.

We built a comprehensive, user-friendly platform that aggregates up-to-date financial data, provides daily analytics for lenders, and offers machine-learning-driven features for portfolio analysis and real-time underwriting, leading to quick and informed lending decisions.

The first step in building the Karus solution was to create the Karus brand. Because we were building a platform focused on trust, quality, accuracy, innovation, efficiency, and transparency, we needed the brand to reflect those values. Karus aims to revolutionize the area of auto-lending, so our approach was to create brand guidelines that highlighted the modernity of the solution in a market that is overwhelmingly outdated.

Our next step was to understand the pain points of lenders to discover what current solutions existed and how we could build a better one. The main pain point we uncovered is that lenders are currently spread across different tools and different platforms on a daily basis. We discovered that they needed one place to keep track of their cash flow and understand how their loans were performing. They needed a platform to use on a daily basis to see an analysis of loan applications, statuses, and outcomes that go beyond the cumbersome Excel sheet or outdated platforms. We focused on building one platform that could be used daily to see all payments, loans, and analytics in order to ease these pain points.

Our initial challenge was to aggregate the data coming from third parties in one place while updating the platform with the latest information. We needed to ensure that users didn’t lose track of payments, so all the scattered data needed to be up-to-date and ready for the user to see. We created a database that keeps everything in sync, allowing loan applications, applicants, and payments to be updated in real-time so that the users know the data in the portal is accurate and presented in an understandable way.

We built the first iteration of the easy-to-use platform and released it in the fall of 2022.

After the platform was built, we focused on secondary sets of users–institutions and dealerships–and their pain points. We continued to iterate and improve the platform with additional features, including “Real Time Underwriting (RTU).” With this specific feature, Karus was able to replace two separate and expensive tools previously used by dealerships, which ultimately reduced ongoing operational costs and allowed the user to make faster, more efficient, and more quality decisions.

After releasing the RTU feature, we immediately began refining its functionality based on customer behavior and feedback. This approach allowed us to fully replace the existing tools our first dealership beta user was using with Karus in just one month. Within the first month of releasing RTU to this customer, they created over 1,100 applications. Out of these, 100 were booked, with an average loan amount of $16,842, culminating in a total loan value exceeding $1.7 million.

After leading the development and launch of the Karus SaaS product and taking it through commercialization, our mission was successfully completed in June 2023. Karus and their in-house team continue to grow the product and business, leading the charge in transforming the auto finance industry.

Karus is a SaaS product that uses ML models for precise risk assessment, helping lenders understand potential payment defaults. With daily reporting and clean visualizations, users are able to have a clear view of loan behaviors, which allows them to make informed lending decisions and strategies. With Karus’ insights, lenders can approve loans to creditworthy borrowers, earning money by approving subprime loans that previously would not have been generated. Lenders can also see accurate assessments in Karus, which leads to safeguarding their investments. Ultimately, Karus’ AI-powered system simplifies the lending process from start to finish, saving users’ time and resources and earning them more revenue than before.

As of today, 73,000 unique applicants get their loan decisions through Karus, and $30M of loans have originated within the platform. $2T in application volume has been underwritten at the point of sale, and $1T in receivables has been underwritten post-origination. To date, Karus has raised $4.7M and continues to change the game in the auto finance industry. We look forward to watching them grow.