Episode
On .NET Live - OpenAI Assistants Orchestrated with Microsoft Semantic Kernel Agents
with Daniel Costea, Luis Quintanilla, Stephen Toub, Frank Boucher
Semantic Kernel is an open-source SDK that lets you easily build AI assistants that can call your existing code. In this week's episode, community MVP Daniel Costea visits to show us how this easy-to-use SDK can create advanced AI-powered interactions, featuring AI imitations of Jack Sparrow, Don Quixote, Shakespeare, and Yoda!
Chapters
- 00:00 - Countdown
- 03:06 - Welcome to On .NET Live
- 04:51 - The Semantic Kernel Journey
- 05:13 - Generative AI/Large Language Models
- 06:44 - OpenAI Assistants
- 14:30 - Semantic Kernel Agents
- 17:27 - Agents Interaction powered by OpenAI Assistants
- 20:28 - Dialog Script (Steps)
- 22:36 - Hierarchical chat agent interaction
- 24:12 - Architecture
- 28:54 - Demo
- 41:45 - Joint chat agent interaction
- 42:35 - Agent with More Agents as Plugins (One Prompt)
- 43:20 - Demo
- 47:40 - Agent with More Agents as Plugins (One Prompt)
- 49:50 - Demo
- 51:16 - Conclusions
- 52:26 - Resources
- 52:54 - Q&A
- 59:11 - Wrap
Recommended resources
Semantic Kernel is an open-source SDK that lets you easily build AI assistants that can call your existing code. In this week's episode, community MVP Daniel Costea visits to show us how this easy-to-use SDK can create advanced AI-powered interactions, featuring AI imitations of Jack Sparrow, Don Quixote, Shakespeare, and Yoda!
Chapters
- 00:00 - Countdown
- 03:06 - Welcome to On .NET Live
- 04:51 - The Semantic Kernel Journey
- 05:13 - Generative AI/Large Language Models
- 06:44 - OpenAI Assistants
- 14:30 - Semantic Kernel Agents
- 17:27 - Agents Interaction powered by OpenAI Assistants
- 20:28 - Dialog Script (Steps)
- 22:36 - Hierarchical chat agent interaction
- 24:12 - Architecture
- 28:54 - Demo
- 41:45 - Joint chat agent interaction
- 42:35 - Agent with More Agents as Plugins (One Prompt)
- 43:20 - Demo
- 47:40 - Agent with More Agents as Plugins (One Prompt)
- 49:50 - Demo
- 51:16 - Conclusions
- 52:26 - Resources
- 52:54 - Q&A
- 59:11 - Wrap
Recommended resources
Have feedback? Submit an issue here.