How Badax AI works
1. You are talking to an AI system
Consult responses are generated by an automated AI language model (hosted via Azure OpenAI / Microsoft Foundry), not by a human clinician in real time.
2. What the model may receive
Depending on your use, the model may receive: your message, recent chat context, your health profile, and retrieved snippets from documents/examinations you stored (RAG).
3. Limits and risks
The assistant can misunderstand labs, miss contraindications, invent plausible-sounding facts, or omit important caveats. It is not a substitute for professional medical judgment.
4. Intended use (draft)
[TO BE COMPLETED BY COUNSEL — placeholder only, not legal advice] State intended use, e.g. “personal information organization and educational-style explanations of materials the user already has,” and explicitly exclude diagnosis/treatment claims if that is your position.
5. Human oversight
[TO BE COMPLETED BY COUNSEL — placeholder only, not legal advice] Describe any human review, support escalation, or reporting channel for harmful outputs: [EMAIL].
6. More information
See also the Privacy Policy (subprocessors, retention) and Terms of Service (liability and acceptable use).