The Federal Housing Finance Agency's second annual tech sprint held from July 22 to July 25 highlighted how conceptual generative artificial intelligence applications can remove certain pain points in single-family and multifamily housing transactions.
Out of the 12 conceptual projects presented, four were picked that were considered the best use cases to streamline four categories including the customer experience, assess creditworthiness, solve risk management and compliance and modernize operations.
Though the applications are theoretical, there is potential that similar applications could come to the forefront in the near future.
Speaking at the beginning of the tech sprint Sandra Thompson, the director of the FHFA, highlighted that last year's tech sprint resulted in a number of innovations that were implemented by Freddie Mac and Fannie Mae.
"One idea advanced the concept of a centralized portal to improve access to down payment assistance programs for potential home buyers," Thompson said. "And there are now several tools, like Freddie Mac's DPA One, which is the down payment assistance that help lenders easily identify programs for homebuyers that they might qualify for."
Thompson added that FHFA, Fannie and Freddie are actively looking at ways to "responsibly innovate."
"The housing finance industry depends largely on paper driven processes, and it's been my way for well over 30 years. By harnessing the power of innovation in our industry, we can also make progress towards solving long standing challenges our industry faces, such as closing the homeownership gap and addressing the lack of credit access and credit availability in underserved communities," she said.
Here is a breakdown of the projects that were winners and how they address challenges in the single family and multifamily space.
Best consumer experience idea
Team Open Angels, led by Melissa Langdale, president of the Mortgage Collaborative, and Niraj Patel, an AI digital transformation executive at Greystone, proposed an application that would help standardize the rental application process. With the help of generative AI, the platform would provide a digital and transparent experience for tenants and property managers. Specifically, the portal could aid those looking to rent a property by analyzing their financial standing and making a judgment of whether they could qualify to rent an apartment, or if not, how they could go about improving their credit. When a tenant is ready to apply for an apartment, only specific documents required by a property would be sent, thereby alleviating potential burdens for property managers.
Best assessing creditworthiness idea
Team dubbed AI-Astronauts, which included members Eric Lapin, president at Freeform, Shanth Ananthuni, vice president of automation and digital transformation at Cenlar, and Meghan Tidgewell, a facilitator at MISMO, introduced an app streamlining some parts of servicing. Specifically, the app byway of generative AI would be able to identify homeowners experiencing financial hardships. The application would alert both the servicer and the borrower of missed payments and help to facilitate solutions to keep a borrower in their home and to maintain their creditworthiness.
Best operations idea
Team FAIR took home the best operations idea accolades. The group built an application focused on fraud prevention in the multifamily space.
Specifically, the app uses generative AI to analyze and summarize a borrower's documents to check for quality control, identify anomalies and recommend action steps for lenders and agencies underwriting a multifamily deal.
Best risk management and compliance idea
Team Arsenal built a tool to help disabled individuals find proper multifamily housing. One of the members explained the issue has been top-of-mind because his close relative, who is disabled, has had issues finding suitable places to live.
The application built by the team identifies potential gaps in the Americans with Disabilities Act (ADA) compliance in multifamily buildings to ensure Fannie Mae, Freddie Mac purchases are ADA compliant.
