AI Transparency
Full transparency on how we use AI—and more importantly, how we don't.
Last updated: July 2026
Our AI Philosophy
AI should support clinical judgment, never replace it.
We believe AI in healthcare carries unique responsibilities. When we evaluate whether to use AI for a given function, we ask three questions:
- Is this a clinical decision? If yes, AI doesn't make it. Period.
- Can we explain why? If AI produces an output, we must be able to show why.
- Does a human review it? AI-generated content requires human approval before action.
Where We Use AI
Documentation Assistance
AI suggests documentation completions based on assessment context.
Human in the loop: Clinician reviews and approves all suggestions before saving.
Billing Code Suggestions
AI suggests billing codes based on documented care.
Human in the loop: Human review required before any claim submission.
Administrative Tasks
Schedule optimization, routing, workload balancing.
Human in the loop: Suggestions only—staff makes final decisions.
Clinical Decision Support: Deterministic Rules
For anything clinical, we use deterministic rules—not generative AI.
What This Means
- Rules written by clinical experts
- If condition A + condition B → alert
- Every alert traceable to a specific rule
- Same inputs always produce same outputs
- No hallucinations possible
What We Avoid
- Large language models making clinical suggestions
- Black-box recommendations
- Non-deterministic clinical outputs
- AI replacing clinical judgment
- Unexplainable alerts or recommendations
Your Data and AI
- Your patient data is never used to train external AI models
- We do not sell or share data with third-party AI providers
- AI features can be disabled at the organization level
- All AI interactions are logged and auditable
- You own your data—always
Questions about our AI approach?
We're happy to discuss our AI philosophy and implementation in detail.