About
I am a psychologist seeking an experienced full-stack developer to build a secure, internal web application designed to reduce the administrative burden of writing psychological evaluation reports.
The application will act as an "AI Medical Scribe." It will use ambient listening to transcribe clinical interviews, accept manually entered psychometric test scores, and use a RAG (Retrieval-Augmented Generation) pipeline to generate a draft report that matches my specific clinical writing style.
What You Need to Know:
You will not be starting entirely from scratch. We have a highly detailed Product Requirements Document (PRD) and a partial existing frontend codebase 14, React 18, TypeScript) that already has AWS Amplify/Cognito initialized and several complex UI components (like a multi-step dynamic interview wizard) ready to be salvaged and repurposed.
Key Features & Workflows to Build:
Ambient Scribe: Integrate a high-accuracy Speech-to-Text API capable of speaker diarization (separating the examiner from the patient/parents) via a standard laptop microphone.
Word-Level Timestamping: This is a crucial feature. The generated draft text must be mapped to the audio transcript so the examiner can click any sentence in the draft and play back that exact moment in the interview for fact-checking.
RAG Document Generation: Build a pipeline using a secure, zero-data-retention LLM (e.g., Anthropic Claude or GPT-4 Enterprise) that references a secure "Style Guide" database of past, anonymized reports to format the new draft.
Feedback Loop: Implement a feature that scrubs Protected Health Information (PHI) from newly finalized reports and adds them back into the Style Guide database to improve future AI generation.
HIPAA Compliance: Ensure end-to-end encryption and secure data handling using AWS infrastructure (S3, Cognito, etc.). You must understand how to handle PHI legally and securely.
Required Tech Stack & Experience:
Frontend: (App Router), React, TypeScript, Redux.
Backend/Infrastructure: AWS (S3, Amplify, Cognito).
AI/Audio Integration: Proven experience integrating robust Speech-to-Text APIs (e.g., Deepgram, AssemblyAI, AWS Transcribe) and LLM APIs.
Security: Deep understanding of HIPAA compliance, data encryption, and secure cloud architecture.
To Apply:
Please include the following in your proposal:
A brief overview of your experience building HIPAA-compliant software or handling sensitive data.
Your experience with audio processing APIs (specifically regarding speaker diarization and word-level timestamping).
An example of a similar application or AI-integration project you have successfully delivered.
Contract duration of 3 to 6 months. with 40 hours per week.
Mandatory skills: Web Application
Languages
- English
Notice for Users
This job comes from a TieTalent partner platform. Click "Apply Now" to submit your application directly on their site.