Human knowledge and feedback are key for building trustworthy digital services, experts say
SDI Productions via Getty Images
By building public assistance systems around the human experience, agencies are more likely to save time and resources, experts say.
As state governments grapple with budget uncertainties and dwindling federal support for modernization efforts, leaders must take cautious steps to ensure they continue building public benefit systems that meet users’ needs for long-term sustainability, a new report says.
A human-centered approach to designing and developing public benefit systems can help agencies do that by better understanding pain points of customers and users that need to be addressed, according to the report released by Nava Public Benefit Corporation.
“Prioritizing user needs isn’t a nice-to-have; it’s foundational to building efficient and effective systems that prevail for generations,” said Kira Leadholm, senior editorial manager of Nava, during a webinar hosted by the organization on Thursday.
The Massachusetts Department of Early Education and Care, for example, is partnering with Nava to modernize its child care financial assistance system to improve families’ access to the benefits, according to the report. The modernized system will leverage cloud-native software and include an internal portal and public-facing application to streamline the process for families and caseworkers.
To do that, EEC is distributing surveys and conducting in-person and virtual interviews with families to identify challenges they face when interacting with the current system, according to the report. Nava and EEC also hold regular meetings to incorporate staff’s feedback on development of the new system, which will help prepare them for operating and maintaining the end product.
Ensuring that government services are built around customers’ experience is particularly critical as the use of artificial intelligence in public-facing systems becomes increasingly prevalent, as many remain concerned about the tech’s trustworthiness and security.
The New Jersey Department of Labor, for instance, has worked to incorporate AI-enabled translation services into its unemployment insurance application system since last year.
A key part of the process was for department staff to work with UI call center agents to help develop plain language translations between English and Spanish terms, said Gillian Gutierrez, senior advisor and director of unemployment insurance modernization.
Spanish-speaking dispatchers could better explain the difference between the terms fired and laid off between the different languages, which was crucial to inform the state’s efforts to improve translations for the UI application, she said.
The department also worked with claimants to test translation outcomes “to make sure we were offering something that was both truthful and meaningful” for situations where, for example, a claimant needed to ask a dispatcher a clarifying question about their application, Gutierrez explained.
Those insights helped the state build a translation glossary, which staff can leverage when using AI translation tools to produce content, she added.
“This process of AI translations puts humans at the front [by building] on their knowledge and capacity,” Gutierrez said, noting that humans are also a key to the end of the process by reviewing translations for accuracy.
Otherwise, “bad translations could end up with bad customer experience [and] more work for the state staff who work on those claims,” she said, which “could show that you don’t take seriously your responsibilities as a government agency.”
To further prioritize human experiences in the state’s efforts to improve service delivery, Gutierrez said the department is also working with Nava to leverage an AI tool aimed at analyzing customer feedback on UI services.
“We’re trying to find ways where technology can give us more clarity and get us to a place where we can increase the trust in the services we offer,” she said.