✨ New Features / 🛠 Improvements
The output template dropdown is always visible when regenerating notes. Even when your organization has only one output template configured, the dropdown now stays on-screen, so you always have a clear view of which template produced your note.
Your summary shows up even when the AI reports "no medical content". Previously, if the automatic medical-relevance check flagged a transcript as empty, the summary tab could come up blank even when a completed note existed. Now, if a finalized note is present, it will render in the UI regardless — so you're never stuck without a draft when one has already been generated.
🔒 Beta Features
🔒 Beta Features
Chat has several exciting updates (beta). When Chat is opened from an appointment with an unsigned note, the assistant can now read the note and apply edits you approve — and as you make edits, it proactively recognizes reusable preferences (terminology, formatting, level of detail) and offers to save them as one-click custom instructions that apply to future note generation. In addition, Chat sessions involving a patient now start with the patient's active allergies, active medications, recent diagnoses, recent visits, recent labs, and active care plans already loaded in context, so the assistant is briefed before your first question. Also, if you try to edit a note that's already been signed, Chat now explains clearly that signed notes can't be edited, instead of a generic "tool unavailable" message. You can also stop a response mid-stream if Chat is going somewhere you don't want, and the threads list now loads progressively instead of fetching every thread at once. Chat can now also suggest conversation starters based on questions you ask often, and when you save one, you can optionally link it to an output preference so clicking that starter always formats the reply the way you like. Preferences themselves can now be given a friendly name so they're easier to recognize in dropdowns.
Chat can now read your patient's structured clinical data. For tenants using the new clinical data backend, Chat can directly query structured clinical measurements (range-of-motion panels, observations, conditions, medications, allergies) and turn the results into charts and summaries without re-reading the SOAP note.
Billing code suggestions (select organizations). For organizations enrolled in the billing code recommender pilot, the system now calls the recommendation service and captures traces for review.
Please contact us if you're interested in trying the Chat beta.