Build a SAFE Agent on Microsoft FoundryUse the SAFE Framework, ACS, and ASSERT to enforce agent boundaries at runtime and verify the complete behavior.24 min read·paulolacerda.hashnode.devAug 21
Answers You Can Trust: Grounding Enterprise Agents with Foundry IQOriginally published on the Microsoft Community Hub on July 15, 2026. Enterprise agents need more than fluent responses. They must answer using information that is authoritative, current, permission-a2 min read·paulolacerda.hashnode.dev1Aug 4
Practice the Hard Call: Real-Time Voice Training with Live Voice PracticeOriginally published on the Microsoft Community Hub on June 29, 2026. Some customer conversations are too important—and too difficult—to practice for the first time with a real customer. Traditional t2 min read·paulolacerda.hashnode.dev1Aug 4
SAFE: Designing Responsible Agentic SystemsOriginally published in Towards AI on March 6, 2026. Agentic systems do more than generate responses. They make decisions, invoke tools, update state, and execute multi-step processes that may affect2 min read·paulolacerda.hashnode.devAug 4
Many Models Training with Hyperparameter OptimizationDemand forecasting often requires more than a single general-purpose machine learning model. In scenarios such as retail, energy consumption, or supply-chain planning, organizations may need to train2 min read·paulolacerda.hashnode.devAug 4
