The Medical Letter Generator is a dedicated module within the Aescuris software. It prepares findings reports, discharge letters, consultation and referral letters from dictated notes, bullet points or structured practice data — via the GDT interface, via import/export or via a simple copy step. Locally on your hardware. No cloud. No data leaves your network. Designed and clinically guided by an experienced general practitioner.
Surveys conducted by professional medical associations have reported the same picture for years: physicians spend a substantial share of their working time on documentation. The medical letter is among the most time-intensive items — while at the same time being highly standardised in structure.
Practices typically report several hours of daily documentation time. The largest block: letter and findings texts.
From dictation through transcription to the correction loop. In smaller practices without dedicated transcription staff, the work tends to shift into the physician's own time budget.
Salutation, history, findings, diagnosis, therapy, closing — the structure of a medical letter is largely standardised. This is precisely where the generator helps prepare the draft.
The Medical Letter Generator is a dedicated interface within the Aescuris software. Data is handed over from your practice management software via established interfaces — the draft is created inside the Aescuris dialogue and transferred via a simple copy step into your usual working environment.
Dictate as usual — or use the dictation feature of the Aescuris server, which handles speech capture locally. Patient data is transferred from your practice management software (PMS) via the GDT interface to the Medical Letter Generator. Alternatively, content can be passed in via import/export or a simple copy step into the dialogue. On request, the transfer happens automatically in the background as soon as you open the relevant patient record.
You select the letter type (e.g. findings report, discharge letter, consultation report, referral letter) and add context or bullet points directly in the Aescuris interface. Tone and level of detail can be adjusted per letter.
On the Aescuris server inside your practice, the language model — tuned for German and English medical terminology — produces the draft within seconds to a few minutes. Fully offline, with no patient data leaving your network.
You review the draft in the review dialogue of the Aescuris interface and transfer it via a simple copy function into your usual working environment (PMS, Word, email). On request, more advanced integrations can be set up — for example a direct write-back into the PMS. The clinical assessment and approval of the letter remain with the physician at all times.
The leverage of the Medical Letter Generator differs depending on the structure of the practice. The following use cases illustrate where the module unfolds its greatest effect.
Two full-time physician seats, an MFA team, no transcription staff. Letters are written in the evening after the practice closes — typically 30 to 45 minutes of additional work per day.
Effect with Aescuris: from dictated bullet points or notes taken during the consultation, the letter draft is prepared right away. The evening correction loop is reduced to a clinical review and approval.
Diagnostic specialties (cardiology, radiology, gastroenterology) generate a findings report per patient. Templates exist, but individualising them to the specific medical history and prior findings takes time.
Effect with Aescuris: the system combines findings data with prior reports and the patient's history and prepares a contextually coherent findings report, which is then only reviewed and released by the transcription team or directly by the physician.
Several physicians work on the same patients. Discharge letters and consultation reports are required, often at the end of a long consulting session — with the corresponding risk of error.
Effect with Aescuris: the letter is prepared from the structured course of the episode, including all documented findings, diagnoses and therapies. Hand-overs between treating physicians become more consistent.
A medical letter is more than a text document. It is clinical communication — it has to be precise in content, legally sound, and at the same time appropriate in tone. AI must not guess here. It has to work from what was actually documented. That has been the guiding principle for the Medical Letter Generator from day one.
The Medical Letter Generator processes patient data of the special category under GDPR Art. 9. Aescuris is therefore consistently designed as an on-premise solution.
The language model, patient data and letter templates remain on a practice server inside your network. No cloud routing, no external inference endpoints.
Because Aescuris does not process the data on your behalf — your practice carries out the processing on its own hardware — there is no need for a separate data processing agreement, a frequent hurdle with cloud-based solutions.
The architecture is designed to align with German medical confidentiality obligations. There is no path through which patient data could reach Aescuris or any third party.
Each generated letter is logged with timestamp, source-data reference and model version — for your medical documentation and for any external audits.
In 2026 we are taking on a limited number of pilot practices. You receive the hardware, the setup and the clinical accompaniment — at no cost during the pilot phase.
