Embedded Files
PRODUCT OVERVIEW
The goal is to implement chatGPT into a product by specifying constraints so that the product can perform a set of specialized tasks as a co-pilot.
The product assists the entire consultation process, from pre-examinations to doctor appointments to tasks after consultation.
PRODUCT ARCHITECTURE
I started with the types of inputs and devices involved in this product, then established the co-pilot's fundamental capabilities based on patient needs.
Each of the product functionalities is based on an Azure AI service. This system integrates multiple input methods, from voice commands to allowing patients to engage through multiple devices. This Co-Pilot module, backed by Azure Conversational Language Understanding (CLU), manages critical tasks like disease follow-up and diagnosis.
DESIGNING PATIENT MONITORING SYSTEM
The users interact with the co-pilot through voice. The system extracts key medical data, including blood pressure, weight, and exercise habits, and compares them against a Medical Knowledge Graph. With those conversation data, the system generates a structured report, assesses health risks, and sends alerts to healthcare providers under certain circumstances.
Page updated
Google Sites
Report abuse