Advanced RAG Systems : Designing Intelligent, Context-Aware AI Systems

This Live Problem Solving session explores how modern AI systems retrieve, reason, and respond using advanced RAG and agentic workflows.

Instead of isolated concepts, you’ll learn how to design end-to-end, production-ready systems from improving retrieval quality to building intelligent agents that validate and refine outputs.

If your current RAG pipeline feels basic or inconsistent, this session will show you what’s missing and how to fix it.

What We’ll Cover

• ✅ Advanced Chunking Strategies: Learn smarter ways to split data for better retrieval quality.
• ✅ Context Enrichment: Improve context flow using post-processing techniques.
• ✅ Hybrid Retrieval (Fusion RAG): Combine vector search with BM25 for stronger results.
• ✅ Query Transformation (HyDE): Turn weak queries into high-performing ones.
• ✅ Reranking for Precision: Filter and prioritize the most relevant results.
• ✅ Agentic RAG Systems: Build intelligent workflows that reason and iterate.
• ✅ RAG Evaluation: Measure quality using relevancy and faithfulness metrics.

Prerequisites

• ✔️ Basic understanding of RAG systems
• ✔️ Working knowledge of Python
• ✔️ Familiarity with LLM APIs
• ✔️ Exposure to LlamaIndex / LangChain (preferred)

Note: This is not a beginner session. If you're new to embeddings or vector databases, the pace may feel fast.