{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/926a88546fec44f4862ff3109d638f5c\" frameborder=\"0\" width=\"1728\" height=\"1296\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1296,"width":1728,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1296,"thumbnail_width":1728,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/926a88546fec44f4862ff3109d638f5c-4da7ca6ead5ff647.gif","duration":335.488,"title":"Clinical Copilot Demo with FHIR Guardrails","description":"Hi all, in this demo of our AgentForge clinical copilot, I show how Dr. Chen logs in with multi factor authentication, then I pull patient data from the FHIR API and generate a brief for Margaret Hayes. I highlight two hard coded guardrails, a Verified tab that compares FHIR output with the LLM output to prevent hallucinations, and a rule to show nothing when no patient data is available. I also enable observability in LangFuse, with 13 API calls so far, no failures, and traceability across steps. No action is requested from viewers."}