State Representation for Agents Navigating Stateful Clinical Applications
Agents need structured state, not screenshots, to navigate clinical software safely.
Tomás Ibáñez-Reyes
Section
Agent Architecture
10 stories in Agent Architecture.
Tool Calling vs Pure Vision in Computer-Use Agents
Hybrid agents need both approaches—each covers failures the other can't fix alone.
Tomás Ibáñez-Reyes
Prompt Engineering Patterns for Clinical Screen-Reading Agents
Well-designed prompts, not raw model power, determine whether clinical agents work reliably.
Priya Subramaniam
Deterministic vs. Probabilistic Planning in Agents
Deterministic systems prove what happened; probabilistic ones report what they think happened.
Tomás Ibáñez-Reyes
Memory and Context Management Across Long Agent Sessions
Why agents lose track of multi-day healthcare workflows and how to fix it.
Priya Subramaniam
Multi-Agent Coordination in Administrative Workflows
Computer-use agents enable workflows that span legacy systems with no APIs between them.
Priya Subramaniam
Grounding Strategies for Screen-Based Agents
Why vision models fail to click the right button on healthcare portals.
Marcus Delacroix
Reasoning Loop Architectures in Agentic Systems
How the PRAO loop separates true AI agents from chatbots.
Seo-Yeon Baek
Action Space Design for GUI Agents
Defining which software moves an agent can make determines whether it succeeds at real work.
Marcus Delacroix
Vision Model Selection for Clinical UI Parsing
Most clinical AI vision models fail on real systems, scoring below 9% on live EHR environments.
Tomás Ibáñez-Reyes