How AI/ML Tools Help State and Local Governments Improve Citizen Experience, Claims Management

These days, it’s not enough for government tools to simply be functional. As government entities face a constituency that’s more digitally savvy than ever before, it’s paramount that agencies prioritize building tools and applications that simplify the experience for both public servants and their constituents. This is because a constituent’s experience of government is often through its services, and if those services are difficult or frustrating to access it not only delays access to crucial services, but it also erodes public trust in government itself. 

Moreover, government agencies come up against the other side of difficult-to-use applications, websites and forms, spending large amounts of resources trying to sort through the deluge of constituent queries and paperwork.

The Biden Administration has made this a priority for government organizations, releasing the “Executive Order on Transforming Federal Customer Experience and Service Delivery to Rebuild Trust in Government” in late 2021. The EO calls for all levels of government to design and deliver services with a “focus on the actual experience of the people whom it is meant to serve,” and to deliver services more “equitably and effectively.” The end game, according to the EO, is to strengthen the democratic process by providing “direct lines of feedback and mechanisms for engaging the American people in the design and improvement of Federal Government programs, processes, and services.”

But how exactly can agencies go about this? Emerging technology, like artificial intelligence, may offer an answer.

NYDOL Taps Google Cloud to Quickly Set Up New Program

If AI seems far-fetched or futuristic for government use, consider that many of the user-friendly applications and websites that consumers use regularly today, like Airbnb, already incorporate the technology.

“These applications have all kinds of AI incorporated into them,” says Google Cloud's Head of State and Local Government for the Midwest Region Brad Hoffman, .

A prime example of how agencies have tapped AI to improve operations is the New York State Department of Labor, which in 2021 was tasked with operationalizing the state’s $2.1 billion Excluded Workers Fund. The fund provides financial relief for residents who lost income during the pandemic and were left out of federal relief programs like unemployment and pandemic benefits. Over 90,000 people applied to the Excluded Workers Fund  in the first month alone.

To ensure application review and processing was done in a timely manner and citizens received the assistance they needed, DOL teamed up with Google Cloud to design a mobile-first application that allowed applicants to apply online, anytime, in 13 different languages, check the status of their application, receive updates via text message or email, upload documents and receive pre-paid payment cards in the mail once approved. Specifically, Google Cloud’s technologies built a scalable cloud design that simplified processing and helped fund administrators meet high volumes of applicants. 

NYDOL tapped Google Cloud’s Document AI, or DocAI, tool to recognize uploaded documents, automatically analyze data elements and ensure claim integrity. Additionally, Google Cloud’s Call Center AI and virtual agents delivered information to applicants in their preferred language, in real time. 

DOL also used Google Cloud’s fraud detection tools to identify and prevent applications that had errors, like duplicate driver’s licenses, for example. It assigned each application a “confidence score” against several fraud indicators so adjudicators could easily process the information and determine which claims might be fraudulent — a real concern given that a minimum of $89 billion of the estimated $896 billion in federal unemployment program funds could be paid improperly. 

To date, New York has approved 130,159 claims and dispensed nearly $2 billion in funds to citizens through the Excluded Workers fund.

Ohio Sorts Through Fraud with Google Cloud

During the early days of the pandemic, many of the programs set up to help those in need found themselves early targets of fraud. In Ohio, according to Hoffman, the state received significantly more applications for unemployment insurance than there were working-age adult citizens — an immediate red flag that many claims were likely to be fraudulent. Ohio needed a way to pinpoint which claims were legitimate, and it needed to do so quickly and cost-effectively. 

Google Cloud, which has the ability to handle unconventional data, offered an answer, says Hoffman. 

“The internet doesn't prepare its data for Google, but Rather Google has to confront the reality of whatever our customers look for.  Big Data had its birth at Google and we have been leaders in making sense of unstructured data since the dawn of the internet,” he notes.

Google Cloud used a similar process to sift through the data and sort the legitimate applications from the erroneous ones. It also deployed claims analysis for unemployment insurance to detect, stop and recover previous fraudulent claims and funds. 

Ultimately, Google Cloud helped Ohio set up an easy-to-use application that locked down its claims process and prevented bad actors from successfully submitting fraudulent applications and stealing funds from those with legitimate needs.”

Wisconsin Tackles Backlog with Google Cloud AI/ML

Ultimately, Google Cloud AI/ML tools help state and local governments serve citizens more efficiently and confidently via user-friendly interfaces. In 2020, for example, the Wisconsin Department of Workforce Development saw a wave of claims filed in response to the pandemic’s economic fallout, and an outdated legacy unemployment insurance infrastructure hosted on a mainframe server to support it all. The state was soon left with a backlog of over 750,000 claims. The volume and complexity of the cases were too much, so Wisconsin sought out Google Cloud. 

The state adopted Google Cloud AI/ML to create predictive analytics based on historical data to shorten adjudication decision-making for UI claims. This helped deploy payments faster to eligible people. The state also put to use data models and confidence scores to analyze backlog data and determine the quickest route to approval securely and accurately. Plus, the department used Google's BigQuery to find out where claims were getting stuck in the process and fix those hold buckets, while identifying fraudulent cases. 

With the help of Google Cloud, Wisconsin confidently paid out legitimate claims, deferred questionable claims to human adjudicators and dismissed clearly bogus claims. By August 2021, the state had disbursed $2 billion in unemployment benefits and cleared its entire 2020 UI backlog. 

Ultimately, Google Cloud is injecting cloud modernization, rapid application development and smart automation to aid in everyday state and local government processes. It’s improving how governments engage their customers with virtual capabilities, while positioning internal teams with AI and ML, reporting and data analytics to ensure program integrity, reduce fraud and provide a simpler, more user-friendly constituent experience.