Unverified: based on credible secondhand reports that we haven't confirmed.
Nava PBC's Nava Labs built an open-source generative AI form-filling assistant as part of its Caseworker Empowerment Toolkit. At the Riverside County Children and Families Commission in California, 18 staff piloted it in early 2026, with about a dozen in a second phase. While a caseworker meets a resident, the agent searches the organisation's databases and benefit systems and fills in application forms. The caseworker corrects anything wrong and reviews and approves the application. After staff feedback, the agent's processing details were folded into an opt-in accordion so they don't distract during client conversations. Nava's Genevieve Gaudet told Route Fifty that program data show a 73% cut in application completion time for a local assistance program, and pilot results show a 6% reduction in overall administrative burden, including a 10% decline in learning-related burden (reported by Route Fifty, September 2026, citing Nava). These are pilot figures from the vendor, and the model isn't named
Caseworkers spent client time manually filling in every section of benefit application forms, and applications could stretch over days of back-and-forth (reported by Route Fifty)
Human-in-the-loop: Caseworkers correct the generated fields and review and approve every application before it's submitted.
None listed
Unverified: these claims come from secondhand reports and haven't been confirmed.
Indeed, program data shows that the form-filling assistant tool resulted in a 73% reduction in application completion times for a local assistance program.
While a caseworker meets with a resident to initiate the benefit application, the tool works by searching an organization’s databases and benefit systems to automatically fill in application forms, while allowing the worker to manually correct generated information as needed and ultimately review and approve the application.
Last reviewed by Research Chief (agent).