Powered by WhisperX, Pyannote Diarization, and multi-step LLM analysis for DSM-5-TR ASD screening.
Select your login type to continue.
System Management
Welcome back to ASD-scribe.
Create your physician profile. Requires admin approval.
Enter the 6-digit code from your Authenticator App.
Scan this QR code with Google Authenticator, Microsoft Authenticator, or Site Authenticator.
Or enter secret manually:
Set a temporary password for . They will be forced to change it on their next login.
Your password has been reset by an administrator. You must set a new permanent password before continuing.
Cancel & LogoutUpload audio to start a new visit
Click to upload .wav or .mp3
Select a visit from the Encounter History to view its report.
ASD-scribe is a local, AI-powered transcription and diagnostic mapping tool designed for bilingual (Arabic/English) clinical environments.
The goal is to simplify the clinical reporting process for autism spectrum disorder (ASD) by automating the most time-consuming parts of the diagnostic workflow.
Powered by WhisperX + Pyannote 3.1 for speaker diarization and Qwen3-4B with multi-step analysis, ASD-scribe strictly adheres to DSM-5-TR criteria while keeping all data local.
All processing happens locally. No data sent to external providers.
Ensure minimal background noise. Position the microphone equidistant between speakers.
.wav and .mp3 files. High-bitrate recommended.
Arabic and English. Automatic code-switching detection.