AI Disclosure

Every proposal a vendor sends to a practice gets a score from 0 to 100 and a band: Strong fit, Good with questions, Proceed with caution, or Not recommended. The score comes with short reasons, flags and questions to ask the vendor.

PracticeStrategy uses its own scoring rubric (rubric v2), which combines fixed rules with AI judgment. A score covers eight areas:

• Value and fit with the practice's goals
• ROI and payback
• Price compared with the practice's budget
• Fee transparency
• Contract terms
• Integration with the practice's EHR
• Workflow burden
• Risk and responsibility

• Fixed rules (the same input always gives the same result) handle what can be checked from what was entered: price against the practice's budget and current vendor costs, fee transparency, contract terms such as cancellation notice and prepayment, EMR integration, and the starting point for ROI. The rules also set every weight, adjustment, cap and band. Weights shift with the kind of practice (for example, a small independent practice or an ACO-aligned group).
• AI judgment, from Claude, a model made by Anthropic, covers what needs reading: value and fit, workflow burden, risk and responsibility, a small credibility adjustment to the ROI claim, and yes/no flags such as vague deliverables, risk shifted onto the physician, or a product that would handle patient data. The rules then apply those flags.

The score uses only what the vendor submitted and what the practice entered in its profile. Nobody checks whether the vendor's claims are true.

PracticeStrategy administrators can see scores to run and fix the scoring service.

• A score is decision support. It is not legal, financial, medical, compliance or purchasing advice, and it is not a guarantee.
• AI can misread a proposal, miss context, or rate a weak proposal highly or a strong one poorly.
• A score is only as good as the information behind it. If the practice profile is incomplete or the vendor's claims are wrong, the score will be too.
• Always read the proposal, ask the vendor questions, and do your own checks before you sign anything.

Ask AI is a chat assistant, also powered by Claude from Anthropic. It sees your own profile and the conversation, and nothing else: not your proposals, scores or other accounts. Its answers can be wrong in the same ways. Each plan has a daily Ask AI limit, shown on the pricing page.

If patient information was entered by mistake, email privacy@practicestrategy.io and we will delete it.

For scoring, the AI receives the receiving practice's profile details (not its name or location) and the text of the proposal. Attachments are never sent to the AI. For Ask AI, it receives your profile and your messages. The Privacy Policy lists the exact fields.

If a score looks wrong or you have questions about how scoring works, email support@practicestrategy.io. See also our Terms of Service.