09-24, 13:20–14:50 (Europe/Amsterdam), Katherine Johnson @ TNW City
Retrieval-Augmented Generation (RAG) systems rely heavily on the quality of the retrieval process to generate accurate and contextually relevant outputs. In this 90-minute tutorial, we explore practical techniques to enhance retrieval across three key stages: pre-retrieval, mid-retrieval, and post-retrieval. Participants will learn how to optimize data preparation, query strategies, reranking, and evaluation to significantly improve the performance of RAG systems. A real-world case study will guide attendees through implementing these methods in a complete retrieval workflow.
The tutorial provides a structured approach to optimizing retrieval in RAG pipelines. We'll cover:
* Pre-retrieval optimization: how to preprocess and structure data, choose appropriate embeddings, and design efficient indexing and caching mechanisms.
* Mid-retrieval optimization: strategies for query expansion, context-aware reformulation, and reranking retrieved documents for relevance.
* Post-retrieval optimization: methods for refining and filtering results, incorporating user feedback, and continuously evaluating system performance.
Participants will explore practical techniques and open-source tools to operationalize enhanced retrieval at each of these stages. We’ll demonstrate the full workflow through a detailed case study, showing how to implement different techniques at each stage to build a more efficient and intelligent retrieval pipeline.
Outline:
- Introduction to Retrieval in RAG Systems
- Case Study Walkthrough: Applying All Stages
- Pre-Retrieval Optimization
- Data preprocessing
- Embedding techniques
- Indexing strategies
- Pre-filtering and caching
- Mid-Retrieval Optimization
- Query reformulation and expansion
- Contextual query reframing
- Recursive summarization frameworks
- Reranking techniques
- Distributed retrieval methods
- Post-Retrieval Optimization
- Result filtering and refinement
- Contextual relevance using chat history
- Feedback loops and active learning
- Evaluation metrics and continuous monitoring
- Q&A and Wrap-up
Background Knowledge Required:
* Familiarity with Python
* Basic understanding of LLMs
Target Audience: ML engineers, data scientists, developers working with LLMs in production, and anyone looking to learn how to build robust AI workflows using open source tools.