Beyond RAG basics: Strategies and best practices for implementing RAG

Maxime Voisin

Senior Product Manager

Cohere

Lily Adler

Principal Solutions Architect

Elastic

Join us to explore advanced techniques in retrieval augmented generation (RAG). This talk is for developers, data scientists, and AI enthusiasts, and provides essential insights to elevate your RAG systems.

If you've played around with RAG, and are looking to optimize existing implementations, our speakers will discuss practical steps to help you build RAG-based systems that run in production and at scale.

We'll take a look at the retrieval phase, covering topics such as: chunking strategies, selecting embedding models, and evaluating performance. We'll also cover topics related to advanced generative techniques, access control, and performance evaluation to ensure that your system works for the right users at a sensible cost.

Register today to learn about the latest patterns and techniques for superior RAG performance!

Highlights

• Building your retrieval pipeline: Strategies for data organization, chunking, and model selection
• Integrating data with generative AI: Seamless data passing, multi-query techniques, and use cases
• Deploying RAG systems: Environment options, total cost of ownership, and interface design
• Introduction to Elastic & Cohere: Overview of powerful tools to enhance RAG

Additional resources

• Elasticsearch open inference API natively supports Cohere's models
• Notebooks on Elastic GitHub for working with Bedrock, Azure OpenAI, Cohere, LlamaIndex, and more