Using LlamaIndex to Build an Open-Source NotebookLM Alternative
summary
Built an open-source NotebookLM alternative called NotebookLlama that uses LlamaIndex Workflows 1.0 and Llama Cloud (Llama par + Llama extract) to parse PDFs with images, tables and varied layouts, extract summaries/highlights/Q&A, generate mind maps and serve a Streamlit UI. Architecture runs as three MCP tools (process file, mind‑map, doc Q&A), workflow is event‑driven and debuggable, components are reusable or runnable locally with a local vector DB, and the demo is already executed and producing results.
AI-written summary · about
chapters (15)
- 0:00NotebookLM alternative intro
- 1:23What is NotebookLM
- 3:33Challenges we faced
- 5:22LlamaIndex tools used
- 6:03LlamaIndex workflows
- 7:29LlamaExtract for document processing
- 9:39Building NotebookLlama
- 12:51Open-source requirements
- 13:55NotebookLlama architecture
- 15:44Workflow design
- 17:10MCP reusable tools
- 19:42Document Q&A with citations
- 21:31Vector DB and model flexibility
- 22:13Serving as MCP servers
- 23:17Call to action and Q&A
related talks
42mUsing LlamaIndex for building Generative AI apps
29mUsing LangChain and LlamaIndex to customize LLMs...
01h 28mBuilding LLM Agents Made Simple a
11mIntroducing Llama API: A Seamless Way to Revolutionize Your Workflow
32mEvaluation of QA systems built on LlamaIndex
25mA Deep Dive Into Retrieval-Augmented Generation with LlamaIndex
36mBuilding & Deploying LLM Apps
49mNext-Gen Documentation with AI