MediDictate AI is a HIPAA-compliant, browser-based speech-to-text transcription tool built for radiology and other clinical specialties. Physicians simply record their findings via the built-in microphone interface, then hit “Process with AI” to convert audio into text, extract key data with an LLM, and auto-populate structured report fields—patient demographics, study details, indications, comparisons, and narrative impressions. Drafts can be saved, edited inline, and finalized in one seamless workflow, dramatically reducing documentation time and improving report accuracy.
Tech Stack Overview
MediDictate AI runs on a Next.js/React front-end styled with Tailwind CSS, talking to Next.js API routes (Node.js + TypeScript). Audio capture in the browser feeds OpenAI Whisper (or optional Deepgram) for real-time STT, then an OpenAI GPT-4 (or interchangeable LLM like Gemini or Gork) pipeline extracts key clinical fields and populates report forms.
Frontend



No items found.
Backend & API


No items found.
AI, LLM, STT & TTS




No items found.



Design
Develop
Standout
Design
Develop
Standout
You got a project in mind?
Let’s collaborate and create a powerful brand for your community