Snowflake-Native Enterprise RAG& Knowledge Retrieval | Snowflake Cortex AI, Cortex Search, CortexAgent, Confluence, Python, SQL

? Designed a Snowflake-native retrieval-augmented generationsolution for internal Confluence knowledge content, converting enterprisedocumentation into an AI-searchable knowledge layer for engineering andoperational users.

? Built the ingestion flow using Confluence REST APIs, prepared andchunked source content for retrieval, generated Cortex AI embeddings, andenabled semantic and hybrid search over enterprise knowledge.

? Integrated Cortex Search with Cortex Agent to ground LLM responsesin governed internal content and provide a conversational experience overtechnical documentation.

? Improved knowledge retrieval from approximately 5-15 minutes ofmanual searching to under 7 seconds while reducing token cost by 93% throughretrieval optimization and Snowflake-native processing.

? Evaluated chunking, retrieval relevance, prompt context, andresponse quality to improve grounding and minimize unnecessary context sent tothe language model.

? Designed the solution to keep enterprise data within Snowflakeservices, supporting security, governance, and simplified operationalmanagement.