n8n Workflow - 2 Worklfows
Budget: €30 – €250 EUR
Hi,
I hope you are doing well.
I am looking for an experienced n8n developer to build a secure and automated workflow for my medical practice. The goal is to create a structured dictation-to-document pipeline.
Here is the intended workflow:
I manually start a dictation using Whisper (local or API-based).
Example spoken input:
"Echo shows known reduced EF of 40 percent and duplex is stable."
The transcript should automatically be saved into a defined input folder (e.g., /dictation_input).
Trigger:
As soon as a new transcript file appears in that folder, n8n should trigger automatically.
Processing Step:
The transcript should be sent to OpenAI (GPT-4 or similar) with a predefined system prompt such as:
"Rewrite the following transcript into a professional medical conclusion suitable for a physician letter."
Expected AI output example:
"Fortunately, the cardiac examination revealed a stable situation in terms of the ultrasound findings. The systolic left ventricular function remains mildly reduced with an ejection fraction of 40%. Duplex sonography also demonstrated stable findings."
Output:
The finalized AI-generated text should then be:
a) saved as a new file in a defined output folder (e.g., /final_letters)
b) optionally formatted as .txt or .docx
c) ideally structured in a way that can later be imported into my practice management system.
The Second automation should be exactly the same but the text file should be split. So everythin regarding the echo text should go into folder A and everything in terms of text regarind the duplex should go in folder B. For example: ChatGpts creates two files
A: "Fortunately, the cardiac examination revealed a stable situation in terms of the ultrasound findings. The systolic left ventricular function remains mildly reduced with an ejection fraction of 40%
Second file B: Duplex sonography also demonstrated stable findings
File A is saved in a different folder then File B. Thank you very much
I hope you are doing well.
I am looking for an experienced n8n developer to build a secure and automated workflow for my medical practice. The goal is to create a structured dictation-to-document pipeline.
Here is the intended workflow:
I manually start a dictation using Whisper (local or API-based).
Example spoken input:
"Echo shows known reduced EF of 40 percent and duplex is stable."
The transcript should automatically be saved into a defined input folder (e.g., /dictation_input).
Trigger:
As soon as a new transcript file appears in that folder, n8n should trigger automatically.
Processing Step:
The transcript should be sent to OpenAI (GPT-4 or similar) with a predefined system prompt such as:
"Rewrite the following transcript into a professional medical conclusion suitable for a physician letter."
Expected AI output example:
"Fortunately, the cardiac examination revealed a stable situation in terms of the ultrasound findings. The systolic left ventricular function remains mildly reduced with an ejection fraction of 40%. Duplex sonography also demonstrated stable findings."
Output:
The finalized AI-generated text should then be:
a) saved as a new file in a defined output folder (e.g., /final_letters)
b) optionally formatted as .txt or .docx
c) ideally structured in a way that can later be imported into my practice management system.
The Second automation should be exactly the same but the text file should be split. So everythin regarding the echo text should go into folder A and everything in terms of text regarind the duplex should go in folder B. For example: ChatGpts creates two files
A: "Fortunately, the cardiac examination revealed a stable situation in terms of the ultrasound findings. The systolic left ventricular function remains mildly reduced with an ejection fraction of 40%
Second file B: Duplex sonography also demonstrated stable findings
File A is saved in a different folder then File B. Thank you very much