Ex-Brex COO and Current Unicorn Fintech Figure Leader Claims GPT is Transforming the Mortgage Industry

Lending startup Figure has announced the launch of innovative AI tools designed to streamline the home lending process. The new feature, powered by GPT-4, aims to identify errors in lending documents, significantly enhancing efficiency.

Founded in 2018, Figure focuses on enabling consumers to secure home equity lines of credit. The company's all-online platform reduces the standard 45-day lending process to just five days. Over 50% of Figure’s operations are business-to-business (B2B), partnering with companies such as solar panel loan provider GoodLeap.

With over $1.5 billion raised and a latest valuation of approximately $3 billion according to PitchBook, Figure is now embracing AI technology under the leadership of new CEO Michael Tannenbaum. Previously the COO at Brex, Tannenbaum believes that AI has the potential to revolutionize operations in fintech. "I thought this could truly transform how fintech businesses operate," he stated regarding his transition.

A key AI initiative led by Tannenbaum focuses on “stare and compare” scenarios within the lending process. He illustrated this by referencing the property-level description—a unique asset depiction that must match across various legal documents. Traditionally, a human would need to manually verify this across over 60 pages. Tannenbaum highlighted that the new feature substantially reduces the manual labor and time involved in document verification, exemplifying how AI can “reduce costs in complex processes.”

Considering the sensitive information involved in loan applications, Figure worked closely with OpenAI to ensure a robust privacy agreement, ensuring that customer data is not improperly used for model training.

While the AI feature utilizes GPT-4, Chief Data Officer Ruben Padron stressed the importance of building model-agnostic systems. “We continuously test and assess new models as they emerge, often on a daily basis,” he explained. Their adaptable systems offer flexibility to swiftly pivot to the vendor delivering top performance.

Padron envisions expanding AI capabilities at Figure, reinforcing that increased automation in the lending application process minimizes errors and biases. "Our goal is to lower costs, eliminate manual tasks, and reduce bias," he said. “It's an ongoing journey, not a fixed destination.”