Orchestrate your AI with Semantic Kernel
summary
Shows how Semantic Kernel wires GPT models and embeddings into reusable pipelines using Azure OpenAI (GPT‑3.5/GPT‑4) and vector stores, demonstrates a live summarizer + transform pipeline that actually runs, and uses retrieval-augmented generation to ground answers from indexed docs. Highlights plugins/skills, a sequential planner that picks relevant skills at runtime, and embedding storage options (Azure Cognitive Search, Redis, in‑memory). Java support lives on a Java alpha branch (examples on GitHub); most demos work, one example fails but planner and Yoda‑style transform succeed.
AI-written summary · about
chapters (15)
- 0:00Orchestrating AI with Semantic Kernel
- 2:13Generative AI background
- 10:09Generative AI limitations
- 11:56Azure OpenAI models overview
- 12:18Prompt engineering for LLMs
- 18:27Traditional AI vs LLM integration
- 20:57Semantic Kernel SDK intro
- 23:07Semantic Kernel setup and example
- 28:09Plugins in Semantic Kernel
- 32:22Skills and function pipelines
- 36:34Planners for dynamic tasks
- 43:06Embeddings and vector databases
- 46:43Retrieval augmented generation
- 51:13Semantic Kernel resources and future
- 53:22Q&A
related talks
33mHow to enhance your own AI services using Semantic Kernel
47mAI Orchestration with Semantic Kernel
46mBuilding AI solutions with Semantic Kernel
01h 00mBuilding your own AI Agent using Semantic Kernel
53mUnleash the power of AI: Automate tasks with OpenAI and Semantic Kernel
01h 02mDevelop AI agents with Semantic Kernel
26mSemantic Kernel: Teaching AI to Not Just Repeat, But Complete!
48mGetting started with generative AI using Azure OpenAI Service