GenAI Stack for Developers
summary
Claims a reproducible GenAI stack that runs Llama2 / GPT-3.5 / GPT-4 locally via Ollama, orchestrates chains and deployment with LangChain (including LangServe and LangSmith), and uses Neo4j knowledge graphs plus hybrid vector search for RAG under Docker. Cites surprising failures—GPT‑4 verifies prime numbers only ~2.4%—and cites benchmarks: graph‑aware embeddings give ~10x hit‑rate improvements over vector‑only search and Microsoft finds graph RAG outperforms baseline. Customers move a graph‑RAG pipeline into production using Neo4j graph data science and hybrid search.
AI-written summary · about
chapters (9)
- 0:00Introduction and context
- 3:12Gen stack overview
- 6:03Docker in the stack
- 8:31AMA local LLM management
- 10:16LangChain framework
- 13:29Neo4j graph database
- 17:47RAG and limitations
- 24:33Graph RAG with Neo4j
- 30:41Use cases and conclusion
related talks
39mA New Stack for the Gen AI Era
01h 50mChatGPT & LLMs: A Practical Starter's Guide to Language Chains
35mBoost LLMs with Event Streaming and Retrieval Augmented Generation
34mEnter the Brave New World of GenAI with Vector Search
15mHarnessing the Power of Generative AI for Intelligent Applications
28mBuilding AI Applications in Go
31mBuild an Agentic RAG system using Langchain, Ollama and Milvus
01h 00mUnleashGPT: Mastering the Implementation of Large Language Models