Building Context-Aware Intelligence Pipelines with Open Source - Sudheesh Kairali & Sachindra Joshi

This talk delves into the importance of context in the era of generative AI, emphasizing that data without context is merely noise. The session features industry experts Sudheesh, an IBM Distinguished Engineer and Master Inventor, alongside Sachin, who illustrate how to convert raw, unstructured data into AI-ready assets through context engineering and semantic understanding. They present a practical architecture utilizing open-source tools such as Docling for document intelligence, ContextForge for metadata enrichment, OpenSearch for semantic search, and Langflow for visual workflow orchestration. Key principles for developing intelligence pipelines are shared, focusing on maintaining context throughout data transformations, leveraging semantic relationships, and adopting open standards for flexibility, making this talk a valuable resource for those building RAG systems, knowledge graphs, or AI agents.