Build a RAG app in minutes using Langflow OpenAI and Azure
summary
Live-codes a production RAG chatbot: ingests a Wikipedia page, splits it into 1,000-character chunks (105 chunks), creates embeddings, stores them in Astra DB vector store, runs a vector search returning the top 4 passages, passes ~1,096 tokens to OpenAI (ChatGPT 3.5) and gets the correct release date (March 1, 2024). Exports Python code for a Flask app, deploys as a container to Azure, and adds observability via LangSmith while editing flows in the Langflow UI.
AI-written summary · about
chapters (8)
- 0:00Introduction to RAG
- 1:26Building the RAG pipeline
- 3:44Data ingestion and vectors
- 6:12Optimization and accuracy
- 8:53Deployment to flask app
- 11:13Deploying to Azure
- 11:57Observability with LangSmith
- 14:04Closing thoughts and Q&A
related talks
44mThe RAG Really Ties the App Together
52mBuild an Intelligent Python App, No Infra Hassles!
41mFrom RAGs to riches: Build an AI document interrogation app in 30 mins
23mBuild intelligent apps with JS: integrating RAG, Azure OpenAI and Langchain.js
51mEvent Driven AI: Supercharging ChatGPT with RAG & LangStream
27mFrom RAGs to riches: Build an AI document interrogation app in 30 mins
42mTake Your RAG To The Max!
25mLangChain with Azure AI: Building LLM powered Applications