AI, Human Judgment, and the Future of Child Welfare: A Conversation with Steven Hintze, Arizona Department of Child Safety

The Arizona Department of Child Safety’s three‑year data modernization effort illustrates how a government agency can replace a legacy, desktop‑only child welfare information system with a cloud‑based, mobile platform. Chief Data and Product Officer Steven Hintze explains that the new system lets caseworkers enter notes, run assessments, and retrieve records from any device, eliminating the four‑inch monitor constraints that once hampered field work. By moving the core CWIS to the cloud, the agency gained real‑time analytics, faster decision‑making, and the flexibility needed to meet both state and federal reporting requirements.

Central to the transformation was the Guardian Insight suite, a modular, app‑centric architecture that mirrors consumer‑grade experiences. Instead of a monolithic “big bang” replacement, each functional area—such as rule‑based disclosures, automated invoicing, or data‑quality checks—was built as an independent app that can be updated without disrupting downstream processes. Hintze’s team embedded human‑centered design from day one, conducting on‑site interviews with caseworkers, supervisors, and even families. This genuine curiosity uncovered pain points like excessive clicks, resulting in a redesign that has already returned roughly 15,000 caseworker hours to direct client interaction.

The measurable business value of the modernization goes beyond efficiency; it directly supports the agency’s mission to protect children and strengthen families. By delivering a user‑friendly interface and actionable analytics, the department can identify risk patterns earlier and allocate resources more strategically. Hintze emphasizes that technology must remain a tool for human judgment, not a replacement, and that continuous feedback loops preserve that balance. For other state and local governments, the Arizona example demonstrates that combining AI‑ready data pipelines, product thinking, and empathetic design can produce scalable, mission‑driven outcomes in the most sensitive public‑service domains.