Ajay Shenoy
Evaluating the "R" in RAG
Join us (Ajay Shenoy and Kalyan KS) for a deep dive into the retrieval component of Retrieval-Augmented Generation (RAG) systems. This session will explore practical techniques and evaluation metrics for building high-quality retrieval pipelines that power effective LLM-based products.
What we'll cover:
- The need for RAG in an LLM-powered product
- Chunking your data for optimal representation
- Vectors and semantic similarity
- The drawbacks of semantic similarity
- Evaluating the retrieval - Precision, recall, MRR, MAP, NDCG
- Reranking to improve retrieval
Who should attend:
• ML engineers building RAG systems
• Researchers evaluating retrieval quality
• Builders focused on production LLM use cases