Founded by Veronica Breene, Vesta Insights is on a mission to transform how financial institutions assess creditworthiness, leveraging ‘Explainable AI’ to make lending more inclusive for future borrowers, whilst reducing risk of arrears for lenders. Having recently worked with Vesta Insights prior to their successful seed funding round, we wanted to discuss their AI process and approach with Veronica.
Things kicked off for Veronica during her PhD, when she signed agreements with Freddie Mac and Fannie Mae, two American housing authorities. These agreements gave her access to a dataset of 50 million mortgages spanning 20 years. Armed with this data, she developed machine learning models that not only predicted potential defaults but also explained the reasoning behind these predictions.
Traditional credit scoring methods often exclude individuals who could repay loans, but don’t fit the rigid parameters of existing models. Vesta Insights’ AI technology aims to change that, increasing access to credit for marginalised populations such as freelancers, immigrants and those with limited credit history.
“Our models can expand the pool of creditworthy applicants by 5-10%, depending on the institution’s rules,” says Veronica. This means more people gaining access to the financial resources they need, fostering economic inclusion and growth.
Explainable AI is at the core of Vesta Insights offering, and is something that Veronica champions passionately. As she puts it, “Explainable AI focuses on the human aspect of artificial intelligence, not just what the machine is saying but why it’s saying it.”
This approach is vital for high-stakes decision making, such as lending. By presenting AI generated explanations to underwriters, Vesta Insights enhances their expertise, rather than replacing them. It’s a ‘human-in-the-loop’ system – meaning it’s designed to complement, not compete with, human judgement.
“Our models don’t just predict whether someone will miss a payment; they explain why this might happen.” These explanations empower banks to make better lending decisions while ensuring fairness for customers who might otherwise be denied credit under traditional scoring systems. In turn, this is helping lenders reduce arrears costs by 20% and boost portfolio performance by 10%.
For those hesitant to get started with AI for credit decisioning, fraud detection offers an alternative entry point for getting started with AI. AI systems excel at identifying anomalies, such as fake bank statements or mismatched barcodes, with far greater accuracy than human analysts.
Veronica sees this as a lower risk use case that could pave the way for broader adoption of AI. “Fraud detection provides significant business impact without the high-stakes of credit decisioning.”
Regardless of the use case though, introducing such transformative technology comes with its challenges. “Banks already have profitable systems in place. Convincing them to adopt new technology requires building trust.” To overcome this, Vesta Insights uses a staged approach:
The company is also exploring ways to harness advancements in generative AI within controlled environments. Veronica is particularly excited about the potential of knowledge graphs. These tools, which link information from diverse sources, could revolutionise how data is combined to predict future trends.
“By integrating open banking data, statistical office information, and institutional data, we can create a clearer picture of financial behaviours and triggers,” she explains.
Ultimately, Vesta Insights’ offering is about more than just AI-based technology, it’s about reshaping the financial landscape to be fairer and more inclusive.
“Explainable AI isn’t just about accuracy; it’s about understanding,” says Veronica. By bridging the gap between AI and human decision making, Vesta Insights is setting a new standard for ethical and effective financial technology.
As Veronica and her team look to future market expansion and new features, their mission remains clear: to empower institutions with tools that drive better, more profitable decisions, by enhancing inclusivity, and an understanding of the people they serve.
Whether you have an idea you want to explore, or you’re not sure where to start, the Greenhouse can help.