Dallas Business Journal: How DHA leveraged data, AI to distribute CARES funds to DFW

When it came to distributing CARES Act rent relief, Myriam Igoufe wanted to ensure the process was equitable and funds were being distributed to the most vulnerable households in Dallas.

Igoufe, vice president of policy and development research, along with her team at DHA Housing Solutions for North Texas, is using data and artificial intelligence to distribute more than $4 million in CARES Act funds to those behind on their rent.

After the first round of rental assistance disbursed through the City of Dallas raised questions around if funding was reaching the appropriate households, partners such as DHA were tasked with reaching the neighborhoods and families at risk of eviction.

“We wanted to be able to be more intentional in terms of the equitable distribution of the funds,” Igoufe said.

To determine the most vulnerable neighborhoods, Igoufe adapted the Centers for Disease Control’s social vulnerability index to fit Dallas’ local conditions. Igoufe considered the number of people living in each neighborhood, whether most housing units were rented or owned and CDC data to allocate portions of the rental assistance funds to each city council district.

“The purpose of that methodology is to be able to account for geographical inequities and vulnerability … Some of the patterns that have been observed historically and recently with the first run of funding for the city … is that we need to make sure that we have resources dedicated to chronically under-resourced neighborhoods,” Igoufe said.

Those who apply for rental assistance through DHA have the ability to fill out a mobile-friendly version of the application and have a choice of three languages. The team created these features to account for lack of internet connection or computers at home and language barriers in the communities they are targeting.

Once an applicant turns in their information, about 50 percent of the approval process — which includes whether the household qualifies based on location and income as well as making sure identification and numbers were submitted correctly — is processed through an artificial intelligence platform. This allows DHA to process more applications faster.

“We understand that time is of the essence here and we know that some families may be on the verge of being evicted,” Igoufe said. “We really wanted the process to be as reflective or transparent as possible.”

Landlords and tenants receive email and text message updates each time there is a change in the application process in an attempt to alleviate concerns for both parties, Igoufe said.

Once applications go through the process, DHA will create a lottery to determine which households receive the funding allotted for each district. Igoufe hopes the lottery-style distribution will further democratize the process for those that may have gotten the information later or have less access to the materials they need to apply.

“I wanted to redesign the system that fits a little better what we know of people’s ability and structural barriers that they may face,” Igoufe said. “If we don’t do it, who will? If you are a part of the system, you’re a system agent. You, by nature, are going to lead systemic outcomes. So we really wanted to shift the way we’re serving a vulnerable population by really being intentional and data-driven.”

To view the original article from the Dallas Business Journal, follow this link.