By: Nikki Zeichner and Amy Perez, July 11, 2023 


Unemployment Insurance agencies receive diverse types of claims data through mail, fax, online submissions, and customer service calls. Processing it involves data entry across different systems and the manual nature of this core administrative work in UI agencies negatively impacts staff, claimants, and employers. It also takes time away from important work like fact checking inconsistent data, processing appeals, and proactively working on improving processes.   

“There’s a lot of data entry that we have to do to ensure that information that comes into one system ends up in all the other systems where it belongs,” one manager told us during a recent conversation. In conversations with other state Unemployment Insurance staff, we heard that state staff spend a significant amount of time on manual administrative work such as stuffing envelopes and mailing notices, printing and scanning letters, manually scheduling hearings, and entering the same data multiple times into different systems. 

For the past year, modernization teams at the Department of Labor (DOL) have been helping states identify opportunities to automate rote, non-discretionary, manual tasks, with the goal of helping them speed up the time that it takes to process claims. Our team has focused on Robotic Process Automation (RPA) as a tool to enable automation because many states consider it a “game changer” that has the potential to save time and improve workflows. One state told us, “RPA has reduced some of our processing times by 70-80%.” Another explained, “RPA has helped us to reduce our backlog by automating more repetitive tasks and re-allocating merit staff to more complex tasks." 

What is Robotic Process Automation (RPA)? 

Robotic Process Automation (RPA) refers to a software solution that makes it easy to build, deploy, and manage virtual robots (also known as “bots”). These bots mimic the actions staff take when interacting with existing UI systems and software. Bots can do things like recognize what’s on a screen, perform data entry, navigate systems, identify and extract data, complete non-discretionary tasks, and perform a wide range of well-defined actions quickly and accurately.  

RPA is used broadly in many industries, including within the government and the public sector. Several state UI agencies are currently using RPA to automate non-discretionary processes within Appeals, Adjudication, fact finding, Benefit Payment Control (BPC), fraud investigations, claim verification, and Employer Services. Implementing RPA technology can reduce or eliminate manual processes performed by staff, allowing staff members to focus on performing more complex tasks that require human intervention or discretionary decision making. RPA has helped states reduce their backlogs and process claims in a timelier manner.  

RPA can help streamline work for states that rely on outdated technology systems. It can also help states improve aspects of more modern IT systems that are not so user friendly. 

RPA use cases

Generally, we saw RPA help states manage four categories of administrative tasks that don’t require discretionary decision making:  

  • Pulling data from one digitized textual source and entering it into another  

  • Creating and sending general notifications such as reminders and fact-finding requests 

  • Assembling packets of information/documents, particularly for appeals 

  • Completing a series of tasks to move a work item to the next step in the process 

Below is a list of more specific, key use cases in which states have successfully used RPA to create efficiencies across different stages of claims. States that are not using RPA might still consider these use cases when automating non-discretionary tasks with other tools.

Adjudication/Fact Finding bot use cases 

  • A bot reviews employer fact-finding requests and conducts data entry tasks to enter the separation reason into the benefits system.  

  • A bot checks all potential databases to see if fact- finding documentation was returned from the claimant and if not, creates a failure to respond notification that lets the claimant know what information is needed and how to provide this information.   

Appeals bot use cases  

  • A bot creates appeals packets that are sent out to the interested parties and the Hearing Officer/Administrative Law Judge prior to the appeal hearing.    

  • A bot keys data elements (name, issue, etc.) from the appeals request and into a system for human validation. Once validated, a bot processes the appeal request and adds the documentation provided by the appellant to the appeals docket.

  • A bot clicks through the system and identifies Hearing Officer/Administrative Law Judge availability to assign and schedule appeals hearings.


  • This bot completes multiple tasks: 1) Checks the work registration database to see if a claimant has registered, 2) sends out reminders if the claimant has not yet registered, and 3) if the claimant does not register in time, issues a notification to the claimant explaining that the failure to register with the workforce within the required time frame results in benefits being stopped until the registration is completed.  

  • A bot captures claimant data points on claims such as address, banking info, wages, payment details, etc., and consolidates the data into one location for criminal investigators. The criminal investigators then analyze the results as part of their investigation to identify potential fraud.     

Claim Verification 

  • This bot completes multiple tasks: 1) Checks the work registration database to see if a claimant has registered, 2) sends out reminders if the claimant has not yet registered, and 3) if the claimant does not register in time, issues a notification to the claimant explaining that the failure to register with the workforce within the required time frame results in benefits being stopped until the registration is completed. 

  • A bot takes scanned letters that have been returned by the post office and (after being validated by human staff) enters a note on the claim of the claimant's address that the post office is returning mail from, including the date of correspondence.  

Customer Service 

  • A bot locates all emails sent to multiple email boxes from the same claimant, aggregates them, and sends one email to the correct unit for review and processing. 

  • A bot processes a list (provided by the staff) that has information about claimants who have accidentally locked their claims and need new PINs sent to them.   

Employer Services 

  • A bot inserts the necessary data entry updates for the employer account to multiple systems and uploads the document into a file repository.   

  • A bot checks data points such as current and previous employer response data, income, or reason for separation to determine if a claim can be resolved as previously, currently, or never adjudicated, an issue can be created for it, or employer chargeability should be addressed. 

Return on Investment 

States consistently tell us that RPA reduces costs and improves customer experience by helping them process backlogs faster. “We saved hundreds of human hours by using bots to organize files and autofill forms,” one state told us. While initially states talk about the culture change that comes with bringing new automation use cases into their workflow, they also talk about the benefits and improved job satisfaction that staff report to them.  

RPA has shown soft and hard savings for states and enhances their capacity in new and innovative ways. We heard from one state that "RPA has helped [our] UI Program realize substantial measurable efficiencies, resulting in over 300,000 person hours saved which equates to roughly $2 million. RPA technology does not serve as a substitute for improvements to primary technology systems but provides agility and scalability that are complementary. In addition to these tangible benefits, RPA technology has changed the way we . . .  think about our work to see more opportunities for automation across the UI Program." 

Other states across the country have noticed similar results.  Another state told us, “We used RPA for several workstreams. We saved hundreds of human-hours by using bots to organize files and autofill forms. We [were] able to leverage RPA technology to sort and process determinations, applications, and other paperwork quickly. The bot handled repetitive, paper-driven tasks more efficiently and accurately, freeing up staff to focus on call volume. The bot also helped our wage investigations by compiling detailed claimant and employer wage information across multiple quarters to timely complete the review.” 

Even states that are on legacy systems from the 1980s have been able to see benefits from using RPA, including the quick turnaround time it takes to stand up an RPA bot. One state that operates on a COBOL system shared, “RPA allowed us to tackle a wide range of processes that wouldn’t have been possible without a solution like RPA. Due to the flexibility of RPA, we were able to train the bots to work in many different systems quickly and complete processes which would have taken months or years to do the traditional way.” 

While the struggles facing the UI program with regards to budget, staffing, and technology remain, RPA has demonstrated that it is a formidable tool for states to adopt and gain efficiencies which ultimately improves the UI program. 

Contact us 

If you’re a state interested in learning more about how to streamline your workflows and process claims more efficiently, contact us by emailing the UI Modernization Team and CCing your Regional Office representatives. We can help you identify tasks that are good candidates for automation and provide resources for implementing RPA (and other automation tools) at your agency. We can also help you access funding to cover associated costs. 

Also, check out our RPA toolkit on WorkforceGPS. This toolkit was created to help states understand what RPA is, learn how to apply RPA in their agency, and share lessons learned from other states.