How ZS built a clinical knowledge repository for semantic search using Amazon OpenSearch Service and Amazon Neptune

This article explains how ZS Associates built a clinical knowledge repository for semantic search using Amazon OpenSearch Service and Amazon Neptune.

Specifically, the article covers:

• Overview of the clinical document search platform
• Solution architecture with a document processing layer and semantic search platform layer
• How Amazon OpenSearch Service was used for full-text and embedding-based semantic search
• How Amazon Neptune was used to build a knowledge graph and integrate with OpenSearch Service
• The use of large language models (LLMs) for entity extraction, summarization, etc.
• Customer benefits like visibility of hidden relationships, staying informed, monitoring adverse events, etc.
• Challenges faced like large data volume and reducing inference time
• Conclusion highlighting the benefits of using OpenSearch Service and Neptune