What you'll learn

GitHub Copilot accelerates development, but to adopt it confidently, and to pass the GH-300 certification, you need to understand exactly how it handles your code and prompts. In this course, GH-300: GitHub Copilot: Understanding GitHub Copilot Data and Architecture, you'll learn to use Copilot from the inside out. First, you'll explore how Copilot uses, transmits, and retains your data, including the differences between code completion and Copilot Chat retention policies. Next, you'll discover how Copilot builds prompts using context gathering, Fill-in-the-middle pre-processing, the 4 prompt principles, and how the proxy filter and toxicity filters protect both inbound prompts and outbound suggestions. Finally, you'll learn the end-to-end code suggestion lifecycle, how the LLM behind Copilot is fine-tuned with LoRA, and the practical limitations of Copilot you must plan around. When you're finished with this course, you'll have the skills and knowledge of GitHub Copilot data handling and architecture needed for the GH-300 certification exam.