Feature Requests

Currently, the "Use patient documents as context" feature is limited to 3 documents per session. For new patients with large referral bundles (10-50 documents like referral letters, test results, screenshots), clinicians must either re-upload documents to each session or manually select only 3. This creates a frustrating and time-consuming workflow. The suggestion is to allow clinicians to select and use all patient-level documents stored in the patient profile as context for a session, eliminating the need to re-upload and making already-uploaded documents easily accessible.

I am experiencing a formatting glitch on the macOS desktop app when utilizing Night Mode (black screen background with white text). The copy-paste feature works completely fine if the computer is not in Dark Mode and Microsoft Word has a standard white background. However, when my MacBook is in Dark Mode and I copy a visit note from the Heidi desktop app and paste it into Word, the text color automatically converts to black. Because Word is also in a dark mode format, this results in invisible text (black text on a black page). While highlighting the text and using Word's "Clear Formatting" function occasionally acts as a workaround, it is inconsistent and disrupts clinical charting workflows. Could the pasteboard/rich text formatting payload be optimized so that text colors adapt properly across dark mode applications without hardcoding black text?

Need an export notes feature that allows for multiple record export and individual patient notes or groups of patients' notes to EHR (HealthBridge) or pc/mac

Customer wants to be able to add logos of their clinic and hospital at the top of exported documents. The default export header options don't allow a good appearance. They tried uploading a Word document and PDF with their desired heading/logo but Heidi doesn't support this functionality. This would improve the professional appearance of exported session documents.

I would (as I think many) really like the facility for adding images to documents: 'snapshots' of CT slices, graphical plots of serial lung function both to help the patient and act as 'aide memoires' for me. Is this within scope?

When asking follow up questions, and clicking on Copy (copied to clipboard), the bolded words paste as asterisks. I made sure the Microsoft word settings have the Bold and italic auto formatting button checked off. For example these words were bold in the answer when asking Heidi, but paste like this: Tx Plan: Iron Supplementation: The workaround is to draft as new document, then copy with formatting but seems less efficient to do it this way.

Allow users to generate operative notes by speaking a key phrase (e.g., 'please create operative note for total hip arthroplasty') within a Heidi session, along with additional details like implant size, complications, and estimated blood loss. This would streamline the process of creating procedure-specific notes without manually selecting templates from a dropdown.

User requests the ability to list patients by the dates they were seen in the clinic, with accurate times corresponding to their area. This would help them complete referral letters accurately by showing the actual date the patient was seen rather than the date the letter is generated. They want this integrated into their view so they can directly navigate to the dates when the clinic happened to generate letters.

Hi Heidi team, Following my chat with Vine on July 5: Heidi's field recognition fails on certain government medical forms (e.g., VAC PEN 6221/6235), detecting 0 fields and reporting the PDF as unsupported or corrupted. The cause: these are XFA forms (XML Forms Architecture, made in Adobe LiveCycle), not standard AcroForms. In XFA files, the form layout, fields, and data live in embedded XML packets rather than as PDF page objects — the only visible page is a "Please wait..." placeholder, which is what Heidi is actually parsing. Standard field extraction therefore finds nothing. This matters because many legacy government and insurance forms are still distributed only in this format, and clinicians must use them as-is. I have about thirty such forms; manually recreating fields for each is not feasible. It's solvable: other AI tools (Claude by Anthropic) now fill these correctly by reading the XFA "template" packet for field definitions and writing values into the "datasets" XML packet. I'd ask that Heidi (1) at least detect XFA forms rather than flagging them as corrupted, and (2) ideally support filling them. Happy to share sample files or test a beta. Best regards, Stefan Eberspaecher