The shift to digitize licensing with the intent of increasing efficiency is gaining ground in government agencies for a variety of processes, from marijuana licensing to permitting for gun ownership. Many agencies have found legacy licensing systems are outdated or overly complex.
For DSPS, the launch of the MavQ AI platform is part of a larger, multi-phase effort to modernize the department’s infrastructure, according to DSPS Secretary Dawn Crim.
Crim said occupational licensing was originally slated to be the third phase of the project, but it was moved forward in the plan to prioritize the needs of the workforce and to provide family-sustaining wages.
“We administer over 244 [different types of] licenses for occupations, and they can’t wait,” Crim said. “The pandemic really showed us that our workforce needs to have the credentials to get working.”
The department is a high-volume agency, issuing nearly 1 million licenses every two years.
Primarily, the new platform will offer the ability to automate data entry, a time-consuming process for DSPS employees. Digitizing this component will allow the DSPS staff to use their expertise to review and process applications, rather than to open mail and manually enter information into the system — the current process.
In addition, the existing troubleshooting methods are complicated when incomplete applications are submitted, requiring employees to call the applicant for the correct information — provided that applicants have included an accurate telephone number.
The platform ultimately aims to reduce the amount of time spent on troubleshooting in the call center, which Crim noted is currently fielding an average of 4,500 calls per week between a team of six people.
Currently, DSPS is in the early stages of this process and working to move the tool forward, as different licenses have varying complexity levels. The goal is having processes active for the top 10 health licenses by the end of the year. The pandemic underlined the importance of essential workers within the health-care workforce, Crim said.
Throughout the next several months, the department will begin examining the impact on licenses and the call center and any necessary adjustments that need to be made.
An important factor in the change management process was ensuring DSPS staff were involved, as they would need to be laying out processes in a uniform way using best practices.
“It’s really important to do that pre-work,” Crim said. “It’s important to acknowledge the [members of the] workforce that are the experts in a process and bring them to the table as part of the planning process.”
After that, collaboration between DSPS and developers has involved an interactive approach, Crim said, walking the team of developers through the step-by-step processes currently in place: from a piece of mail being delivered to completing the processing of an application.
The features that were most important to DSPS included building a platform with the ability to recognize license type, automate workflow processes and improve efficiency, MavQ cofounder and chief operating officer Vamshi Vaddiraja told Government Technology in a written response.
“Supervised, unsupervised and reinforcement training of the models are in place to continue to evolve with the emerging needs of the agency,” Vaddiraja added.
The AI component adds speed, efficiency and predictability, according to Vaddiraja, which offer the state’s residents a better experience as well as measurable outcomes.
Brent Mitchell, the managing director for U.S. state and local and Canada public sector with Google Cloud, said in a written statement that Google Cloud AI technologies built into the platform enable the state to automate processes like reading emails and extracting metadata from attachments. This allows the state to efficiently bring licensed professionals into the workforce — from doctors to tattoo artists, he said.
