Scaling AI agents in regulated environments requires balancing automation with strict requirements for auditability, consistency, and human oversight. In this session, SAS speakers Joshua, Diana, and David introduce a “Human Approval Loop (HAL)” framework that uses similarity search and decision orchestration to intelligently route only high‑risk or novel requests to human review. You’ll see how SAS Intelligent Decisioning enables this approach by combining rules, models, and LLM-driven agents into a governed workflow that learns from prior decisions over time. The result is a more scalable, explainable, and consistent human‑in‑the‑loop system for enterprise AI.
Presenters: Joshua Beck, Diana Maris, and David Weik, SAS
Watch the recording
