Ambient documentation across 12 senior living facilities8.4 hours per nurse recovered every week. HIPAA-grade.
The operational reality.
FL senior-living group, 12 facilities, 1,800 residents. Nursing turnover at 28% YoY; charge nurses spending 3.5 hrs/shift on documentation; MDS submissions consistently late; survey-readiness costing weekend overtime.
What we ship.
Ambient scribe agent runs on shift devices (mobile + nurse station) capturing care interactions in real time, drafts notes in PointClickCare format with coding suggestions, and surfaces anomalies to charge nurses. Compliance agent pre-fills MDS and survey docs from clinical notes. Family-communication agent handles routine updates with HIPAA-grade audit logs.
The agent topology + Microsoft AI stack.
Agents (Microsoft Agent Framework 1.0)
- Ambient scribe
- Charge-nurse anomaly
- MDS pre-fill
- Family communication
Models
- GPT-5.5 (notes)
- MAI-Transcribe-1 (speech)
- MAI-Voice-1 (family)
Tools (MCP)
- PointClickCare (FHIR)
- Azure AI Search (clinical KB)
- Survey-readiness store
Governance
- BAA + Purview DLP
- Entra Agent ID per role
- PHI redaction at edge
- OWASP agentic Top 10
What we engineer toward.
Recovered per clinician
8.4 hrs/wk
Reduction in prior-auth cycle time
42%
PHI training-data exposure
0
MDS submissions on time at month 3
100%
What we use.
Other patterns we ship.
Reference scenario. Architecture is the one we ship today; metrics are directional benchmarks drawn from industry data and Microsoft customer reference outcomes. Named, attributed customer stories will be published as engagements mature.
Pattern look like yours?
Tell us the pain and we map it to the right reference architecture. End-to-end proposal back in 5 business days.