The tool, developed by Resultant and launched last November, aims to support job seekers in making career-advancing decisions by leveraging wage record data available to DWD.
The state has wage record data that helps track career progression, DWD Chief of Staff Josh Richardson said. The goal for Pivot was to make that information visible so that people seeking work could see current and future job opportunities.
“Our job is just to make it as easy as possible [for job seekers] to figure out what they’re supposed to do,” Richardson said.
The workforce recommendation engine aggregates disparate information, Resultant President John Roach said, to give residents context and bridge an information gap in the job market. DWD runs the state unemployment insurance program and fosters workforce development, and the two overlap considerably. Resultant’s aim, Roach said, was to help align the functions.
Where does AI fit into the equation? Roach said AI has two primary functions in Pivot. First, based on past and present job market data, AI can project future demand in a particular sector. And second, AI can compile job seekers’ information — from education to work history — and align skillsets with relevant skilling programs to help people advance.
Before AI, DWD had published a list of the “Hoosier Hot 50 Jobs,” a resource that let people see the aggregate list, at a regional level, of in-demand jobs in the state. But, said Richardson, “those lists were very one-size-fits-all.”
AI can offer residents personalized suggestions based on the successful experiences of those in similar situations, which align with job market demand. And instead of offering information on a theoretical basis, Richardson said the technology enables the analysis of massive amounts of information to render data-informed recommendations — and suggested career paths that have proven to be successful.
The rise of AI is expected to play a significant role in workforce development, from its potential to reduce bias in hiring to its potential to transform communication. DWD has leveraged AI before, through the Hoosier Talent Network, but that sunsetted in February 2023 amid a service consolidation, DWD Director of Communications Randy Speith said via email last year.
A key difference between that platform and Pivot, Richardson said, is the latter’s location within the state’s unemployment insurance system — ensuring a steady supply of potential users who have already registered and are known to the state.
“We’ve added no additional steps to that process,” Richardson said. “We’re simply using the information that they’ve already given to us to generate recommendations for them.” The state’s previous data work, shaped in large part by the Management Performance Hub, is what makes it possible for the state to rapidly innovate with technologies like AI, he added.
The tool’s long-term impact remains to be seen, but Richardson said DWD will be exploring several things around its use. First, it will look at how frequently people engage with Pivot and when, after a job loss, they use it. Second, the tool includes a feedback mechanism to let users share whether a recommended occupation is of interest to them. If there’s a lack of interest in a high-demand, high-wage job, Richardson said that information can help employers move to reshape perceptions.
Pivot’s current function is to help a user choose an occupation, but Richardson said its next phase is expected to launch this fall: using AI to offer users recommendations on training providers and making the career advancement process even more seamless.
Longer-term still, the state would like to make Pivot available outside the unemployment insurance system, to support people in other stages of job seeking. And, Roach said, the level of information already in government hands, in state longitudinal data systems, with data from preschool through 20th grade, would make it easier for state and local governments to activate the tool.
“I hope what every state does is they think creatively about ways that they can leverage their data, in this domain and others, to more effectively serve citizens,” he said.