California’s health insurance marketplace further expands AI for document verification
boonchai wedmakawand via Getty Images
By
Chris Teale
Covered California said the tech will now be used for more than two dozen document types to speed up checking eligibility while reducing fraud, waste and abuse at the same time.
LAS VEGAS — California’s health insurance marketplace announced last week it has expanded its use of artificial intelligence to scan documents and eligibility.
Covered California said it would expand its partnership with Google Public Sector and Deloitte to automate document verification for its California Healthcare Eligibility, Enrollment and Retention System, known as CalHEERS.
Using the Google Cloud Document AI tool, CalHEERS will check 25 new and different document types including income and residency status, and verify them in a matter of seconds. Previously, it could take between 72 hours and three weeks for final verification, and it sometimes meant follow-up calls to residents who thought they were already covered.
In a statement released by Google, Covered California Chief Information Officer Kevin Cornish said the marketplace has “transformed the eligibility journey into a real-time, consumer-friendly experience.”
“This advancement enables our staff to focus their expertise on high-value engagement and problem-solving with applicants, ensuring they receive the support they need," Cornish continued.
Covered California has been experimenting with and piloting AI for several years as it looks to speed up eligibility verification, automate enrollment and make the process more efficient for its employees and customers while also modernizing its legacy technology. Already, those pilots have found a 40% reduction in manual tasks, Covered California said, while the AI can also help detect fraud, waste and abuse at the same time.
States have come under fire in recent times for the prevalence of alleged fraud, especially in social service programs, and some have previously called for a more proactive, tech-driven approach to drive it down. Doing that while improving the user experience at the same time shows two simultaneous and effective uses of AI, experts said.
“Regardless of what administration you're working inside of, and whether or not your primary focus is on constituent services or the costs of the fraud, waste and abuse in the system, for most states that we talked about, it's both,” Chris Hein, field chief technology officer at Google Public Sector, told Route Fifty at last week’s Google Cloud Next conference in Las Vegas. “You have to have both things going hand in hand. This is a moment that helps us to say that it can be done in a way that utilizes AI at the forefront.”
Many other states are embracing AI for benefits eligibility and already seeing reductions in their backlogs, faster eligibility verification and improved customer satisfaction. While many are still battling legacy technology, it may be tempting to consider throwing those systems out entirely and replacing them with AI, but that can create instability.
Matthew Schneider, Google Public Sector’s go-to-market leader for U.S. state, local and education, said efforts like those in California are not about AI “coming and fundamentally changing” a state’s infrastructure. Instead, it is about building on what is already there and “providing tremendous operational execution for the agency.”
“If you think about the constituent who's making that application, that's a hard day when you're going and asking for government services,” Schneider told Route Fifty at Google Cloud Next. “For a lot of people, you've now given them the ability to understand [if they are] eligible right out of the gate. You're cutting down that call that has to be made that may have taken weeks to get back. That call was an additional time burden on the state to process these things, but also to stay in communication with the constituent.”
Executives now want to replicate their efforts in California in other states, but given the vagaries of health care policy and compliance regulations in every state, that could be an arduous task. However, they said they are confident of success as the technology is proving itself.
“We have to be able to marry our technology to make sense of where the differences really matter,” Karen Dahut, Google Public Sector’s CEO, told Route Fifty at Google Cloud Next. “It's not as easy as rinse and repeat, although we would love it to be, but we can use the technology and agents to say we did it this way in California, given Iowa's policies, given Illinois policies, what would we have to shape and do differently? We're using AI to help us develop solutions for different states, which is accelerating our ability to go state to state.”
NEXT STORY: Maine governor vetoes landmark data center ban