Location>code7788 >text

Course Online: AI Programming | Semantic Kernel From Beginner to Mastery

Popularity：519 ℃/2025-04-09 11:48:39

Course Plan

This course is designed to be no less than 50 lessons, and 44 lessons have been recorded so far. It has been released and launched on March 15, 2025. Unfinished lessons will be recorded and released as soon as possible. Click the link below to try and buy online:

B Station Class | Semantic Kernel From Beginner to Mastery
B Station Class | Semantic Kernel From Beginner to Mastery
B Station Class | Semantic Kernel From Beginner to Mastery

In order to give back to the .NET community, we specially extracted the first chapter of the course (introduction to basics) for 8 lessons to watch for free:

Course positioning

AI programming course for .NET developers

Course Introduction

From the hard-to-find key when OpenAI was released, to the domestic big model DeepSeek's first-year king, AI's capabilities are refreshing their cognition every day. It is clearly foreseeable that AI will become more and more powerful, but as a developer, in addition to paying attention to the latest technical trends, we need to have the courage to end up doing things and think about it:How can these disruptive technologies be truly implemented into the business to drive integration of existing businesses with AI or innovation in new businesses?

But how do you get started as a .NET developer? Most AI programming courses on the market focus on the Python ecosystem, but Semantic Kernel about the .NET ecosystem can only be based on official documents and a few blog posts. With the frequent and continuous updates of Semantic Kernel, most documents and blogs cannot be kept in sync.

Therefore, in order to help .NET developers gain a foothold in the AI ​​era, this course "AI Programming | Semantic Kernel From Beginner to Mastery".

This course will take you to avoid three major misunderstandings:

  1. There is no need to delve into AI theory: skip complex concepts such as deep learning, model quantization, and neural networks, and directly hit SK's core modules of Prompts, plugins, planners, agents, and process.
  2. Zero environment configuration anxiety: Each lesson comes with a Polyglot Notebook that is out of the box, supporting VS Code, and completing the entire process of "code download → run → debugging" in 5 minutes;
  3. Frontier of dynamic tracking technology: 40 lessons are continuously updated, covering the latest SK official features to help you always get ahead!

Course structure

  1. Basics (Chapter 1-3): From environmental configuration to core functions, consolidate the development foundation.
  2. Advanced chapter (Chapter 4-8): In-depth Agent framework, RAG enhancement, Process workflow, MCP, and unlock complex scenario development.
  3. Practical Chapter (Chapter 9): Through the "eShopSupport" case, the entire process technology stack is connected to restore the real development scenario.
  4. Extension (Appendix): Covers enterprise-level needs such as domestic model adaptation, localized deployment, security protection, etc., and keeps pace with technological evolution.

Course Highlights

  1. No need to worry about environmental issues, just install the .NET9 + VS Code + Polyglot Notebook plugin to get started quickly
  2. From theory to practice, interactive Polyglot Notebook is available every lesson, downloadable and executed without Copy code
  3. After learning the course, you will immediately have your own special Semantic Kernel knowledge base and check it as you like
  4. Free Azure OpenAI API Key for course learning
  5. Course content is continuously updated to ensure that Semantic Kernel updates are synchronized

You will get

🌟 Core skills:

  1. Master the Prompt project and flexibly use multi-template engines to optimize the interaction effect of large-scale models.
  2. Master the core modules of the Semantic Kernel framework and proficient in developing AI plug-ins, agents, and RAG retrieval enhancement applications.
  3. Implement multi-model mixed call, function call (Function Calling) behavior control and log tracking.
  4. Build an AI workflow (Process Framework), complete condition control, step reuse and multi-agent collaboration.

🔧 Practical ability:

  1. Connect with domestic big models and localized deployment (such as Ollama), breaking through technical limitations.
  2. Develop enterprise-level AI applications, integrate engineering capabilities such as logging, security, and dependency injection.
  3. Extend the LLM capability boundaries by quickly converting existing OpenAPIs into plug-ins.

🚀 Frontier Vision:

  1. Deeply understand the AI ​​Agent design concepts and collaboration models, and master cutting-edge technologies such as AgentChat.
  2. Unlock the practical application of semantic search (Qdrant) and text search (Bing) in RAG.
  3. Explore the unified base and MCP (Model Context Protocol) protocol to grasp industry trends.

Course outline

Opening: What's Semantic Kernel?

Chapter 1: Quick Start | Quick Start

  1. Semantic Kernel Get started quickly
  2. Docking with domestic big models
  3. Using a proxy
  4. Integrate multiple large models
  5. Dependency injection
  6. Integration log
  7. Packaging abstraction

Chapter 2: Prompt | Prompt Word

  1. Prompt project
  2. Management prompt words
  3. Prompt word template | semantic-kernel
  4. Prompt word template | handlebars
  5. Prompt word template | liquid

Chapter 3: Plugin | Plugin

  1. Understand Function Calling
  2. Use Plugin to expand the LLM capability boundaries
  3. Function Calling Behavior Control
  4. Kernel Function Filter
  5. Function Calling Log Tracking
  6. Let LLM understand Function Calling Return Value
  7. Let LLM return the Json structure as needed
  8. OpenAPI is a plugin
  9. OpenAPI Plugin Advanced

Chapter 4: Planner | Planner

  1. Organize your Plugin with Planner
  2. Stepwise Planner
  3. Handlebars Planner

Chapter 5: Agent Framework | Agent Development Framework

  1. What is an AI Agent?
  2. Chat Completion Agent
  3. OpenAI Assistant Agent
  4. AgentChat | Agent Collaboration

Chapter 6: RAG | Search Enhanced Generation

  1. What is RAG?
  2. Implement online search
  3. What is Embedding?
  4. Implement semantic search
  5. KM | What's Kernel Memory?
  6. KM | Quick Start
  7. KM | Intake and Retrieval
  8. KM | Built-in ingestion pipeline
  9. KM | Custom ingestion pipeline
  10. KM | Custom chunking strategy
  11. KM | Integrate with SK in plug-in form

Chapter 7: Process Framework | Workflow Framework

  1. What is Process?
  2. Process implements conditional control
  3. Process implementation step reuse
  4. Process United Agent

Chapter 8: Interpretation of eShopSupport Case

  1. Start eShopSupport
  2. Interpretation of eShopSupport

Chapter 9: MCP | Model Context Protocol

  1. What's MCP?
  2. Play with Cline
  3. Use the official C# SDK to play MCP Client
  4. Use the official C# SDK to play MCP Server

appendix

  1. | Unified AI docking
  2. Running large models locally based on Ollama
  3. Prompt word safety
  4. To be continued, continue to be updated
<Previous Next>