Keywords
AI Assistant, Clinical Decision Support

Technologies
Medical LLM, Faster-Whisper, OpenMRS, React

Healthcare practitioners in high-pressure clinical environments face significant cognitive and administrative burdens, as they must divide their attention between patient interaction, information retrieval, and documentation across multiple systems. Tasks such as drafting referral letters and issuing prescriptions after consultations further increase this workload. This fragmentation disrupts workflow and can negatively affect doctor–patient engagement.

This project addresses these challenges through the design and development of a Medical AI-based Assistant (MAIA) platform. The proof-of-concept system integrates real-time speech-to-text transcription with a medical Large Language Model (LLM) to provide context-aware processing, recommendations, and support for clinical workflows. It enables practitioners to capture and analyse consultations, summarise patient medical history, and generate structured clinical documentation, such as referral letters and consultation reports, within a single interface. By consolidating these functions, the system aims to reduce administrative cognitive load, support more focused patient care, and may assist clinicians in identifying rare diseases.