Content Type
Exercise
Duration
20-45m
Level
Introductory
Practice vectorizing and storing textual data using InterSystems SQL, then perform vector searches on the data to quickly and easily find entries that are semantically similar to your search query. Optionally, continue the exercise to use this vectorized data as the source for a simple Retrieval-Augmented Generation (RAG) chatbot application that can answer questions about your data.
Learning Objectives
- Vectorize data to be stored as vector embeddings.
- Run vector searches using InterSystems IRIS® data platform.
- Index vector data using a Hierarchical Navigable Small World (HNSW) index to optimize vector search results.
Optional:
- Build a simple generative AI chat application using Streamlit.
- Use LangChain to vectorize data and run vector searches.
- Integrate vectorized data into a RAG chatbot application.
Prerequisites
- Get an introduction to Retrieval-Augmented Generation (video, 3m).