Building AI Agent Workflows with Semantic Kernel
Authored by Microsoft Developer, this video covers building practical AI agents and workflows with Semantic Kernel and Azure AI Foundry, including a demo of SemantiClip.
Summary
In this episode, Microsoft Developer presents a guide to building interoperable AI agent workflows using the open-source Semantic Kernel project. The focus is on helping viewers practically implement AI-driven solutions, leveraging the agent and process frameworks provided by Semantic Kernel.
Key Highlights
- Demonstration of SemantiClip:
- An open-source AI agent that converts videos into structured blog content.
- Utilizes Semantic Kernel in combination with Azure AI Foundry.
- Source code is available for viewers to explore and adapt.
- Semantic Kernel Overview:
- Explanation of agent and process frameworks available within Semantic Kernel.
- Guidance on how to architect interoperable, reusable AI agents for various workflow needs.
- Orchestration Patterns:
- Insights from Evan Mattson on the different orchestration models supported by Semantic Kernel.
- Guidance on choosing the appropriate framework for different use cases.
- Real-World Application Guidance:
- Step-by-step strategies to build functional AI experiences.
- Discussion about practical concerns and best practices in deploying AI agents and workflows.
Resources
- Semantic Kernel Project: GitHub repository
- SemanticClip Source: SemantiClip on GitHub
- Get Involved: Opportunities to connect, contribute to the open-source ecosystem, or submit your own OSS project to Open at Microsoft.
Connect with Hosts
Series Info
- New episodes every Tuesday.
- Open at Microsoft playlist and project submission info included in the video resources.
This episode is a practical resource for developers seeking to build real-world AI workflows using Microsoft’s Semantic Kernel and related Azure AI solutions.