> For the complete documentation index, see [llms.txt](https://docs.mdintegrations.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.mdintegrations.com/partner/exploring-features/ai-features.md). # AI Features {% hint style="info" %} MDI builds AI into the tools your physicians, support staff, and Automations already use — not into a patient-facing chatbot. These are assistive features: they help your team work faster and help your workflows make lightweight decisions, with a person (or a configured Automation) still in control of the outcome. {% endhint %} ## In the Clinicians App and messaging tools When a physician or support staff member is composing or reviewing a message, a handful of AI-assist features are available right there, inline: * **Proofread** — clean up grammar and phrasing before sending. * **Adjust tone** — rephrase a message to sound more formal, more casual, or more empathetic. * **Translate** — convert a message into another language for the patient. * **Summarize a conversation** — collapse a long message thread into a short recap, useful when a new staff member picks up a case mid-conversation. * **Suggest an auto-reply** — generate a draft reply based on the incoming message, which staff can send as-is or edit first. These live directly in the messaging composer, and each one is opt-in per use — nothing sends automatically just because a suggestion exists. A person still has to choose to use it. ## Questionnaire summaries for physicians When a patient completes a Questionnaire as part of an Encounter, MDI can generate an AI summary of their answers. The physician sees this summary first, which gives them a fast read on the intake before they dive into the raw, question-by-question answers underneath. This doesn't replace the underlying data — it's a starting point meant to speed up the initial review, and it's especially useful on Encounters with long or detailed intake forms where scanning every answer individually would otherwise slow the physician down. ## AI-processing steps in Automations Inside the Automations builder, one of the available Step types is an AI-processing step. Partners can drop it into a workflow to have MDI classify or generate content as part of the automated flow — for example: * Auto-triage an inbound patient message (e.g., "billing question" vs. "clinical question") and branch the Automation accordingly. * Auto-categorize a message or Encounter before routing it to the right queue or clinician. The AI step's output feeds directly into the rest of the Automation — a branch condition, a Tag, a routing decision — so triage that used to require a human glance can now happen automatically, before a person ever sees the item. See Automations for how Triggers and Steps compose more generally, including a full message-triage recipe. ## How it works Across all three of these cases, the pattern is the same: the AI feature operates on data already inside MDI — a message, a Questionnaire's answers, an inbound event — and produces a suggestion or classification for a person or a workflow to act on, rather than taking action on its own. None of these features are exposed to patients directly. The patient never sees "AI-generated" content unless a staff member has reviewed and sent it, in the case of messages, or a physician has read a summary internally, in the case of Questionnaires. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.mdintegrations.com/partner/exploring-features/ai-features.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.