Using Knowledge Graphs in Advanced Domain-Centric RAG Implementation

In this session, we'll explore Knowledge Graphs and their impact on Retrieval-Augmented Generation (RAG) models. AI Engineers will gain expertise in integrating structured knowledge into generative models to develop more precise and contextually appropriate AI systems. Participants will learn to configure graph databases, construct and execute queries on Knowledge Graphs, and optimize RAG systems for various AI applications.

Contents:

• Introduction to Knowledge Graphs and their importance in AI.
• Setting up and configuring a graph database (e.g., Nebula, Neptune, Neo4j, Cosmos DB).
• Constructing a Knowledge Graph using the Llama index and real-world data (demo with Wikipedia data on encryption).
• Utilizing Knowledge Graphs in RAG: Querying Knowledge Graphs for answer generation.
• Integrating queries into RAG for enhanced contextual answers.
• Demo on querying and generating responses.
• Advanced Integration Techniques: Combining Knowledge Graph RAG with Vector Database RAG.
• Custom retrievals to improve generative AI results.
• Demo on configuring personalized retrievers and response synthesizers.
• Q&A

Sunila Gollapudi, with 23 years of experience, excels in strategic technology vision and data-driven platforms. As a Product & Platform Architect, she has launched global products and led technology transformations. She is skilled in data strategy, engineering, and analytics.

A hands-on Cloud Architect, Sunila drives cloud adoption in hybrid environments and specializes in data virtualization and advanced AI with Enterprise Knowledge Graphs. She mentors specialists, teaches advanced AI and data engineering, and fosters innovation.

Author of several books on Big Data and AI, Sunila has worked with Broadridge, Accenture, Oracle, and Royal Bank of Scotland. Currently, she is the Executive Director at JP Morgan Chase and a visiting faculty at ISB.