Short Project Description

Timely, evidence-based guidance is crucial in physiotherapy to reduce waiting times and prevent worsening conditions. This project develops a Digital Expert Twin – a Retrieval-Augmented Generation (RAG) system powered by large language models – to support both patient communication and clinician decision-making. The solution comprises two AI “experts”: a Patient Communication Expert for structured first-line assessment and self-management advice, and a Physiotherapist Sparring Partner that synthesizes the latest research, guidelines, and curricular content with transparent citations.

In close collaboration with physiotherapy researchers and clinical stakeholders, the project will curate high-quality domain sources, design safe and explainable RAG pipelines, and evaluate usability, fidelity, and clinical relevance through mixed-methods studies. Building on pilots with the instruction-tuned NorMistral-7B model, the work aims to deliver a validated prototype that enhances access and quality in physiotherapy, informs education across the full curriculum, and provides a transferable framework for responsible LLM deployment in other health domains.