Hack Together: The Microsoft Fabric Global AI Hack
MIT License
87
stars
24
forks
source link
Project: Ingesting Multilingual Clinical Notes and leverage Azure Fabric + Azure Open AI + Azure CosmosDb MongoDb vCore to create summary and embeddings #53
Disclaimer : This is just a Proof-of-Concept.
This project aims to creates a robust and scalable data processing pipeline that ingests healthcare clinical notes in HL7 format, processes them through multilingual normalization, and leverages the advanced natural language processing capabilities of GPT-4 to generate summaries and embeddings. The embedding can be further used for different RAG application usage. This pipeline aims to enhance the accessibility, usability, and analysis of clinical data across different languages for better healthcare outcomes.
Project aim to show how we can leverage Fabrics Data flow, Lakehouse, Azure Open AI integration and persist the data to Azure CosmosDB MongoDB vCore.
Project name
Ingesting Multilingual Clinical Notes
Description
Disclaimer : This is just a Proof-of-Concept. This project aims to creates a robust and scalable data processing pipeline that ingests healthcare clinical notes in HL7 format, processes them through multilingual normalization, and leverages the advanced natural language processing capabilities of GPT-4 to generate summaries and embeddings. The embedding can be further used for different RAG application usage. This pipeline aims to enhance the accessibility, usability, and analysis of clinical data across different languages for better healthcare outcomes.
Project aim to show how we can leverage Fabrics Data flow, Lakehouse, Azure Open AI integration and persist the data to Azure CosmosDB MongoDB vCore.
Project Repository URL
https://github.com/01graut/Hack-Together-Fabric-AI
Project video
https://github.com/01graut/Hack-Together-Fabric-AI/tree/main/video
Team members
Tanushree123-hub, 01graut