About This Course

This course teaches you to build reliable AI systems with large language models. You write modern, typed Python for AI workloads. You treat prompt design as an engineering discipline, with repeatable methods. You learn to control model output with schemas. You learn how tokens, context windows, and inference parameters shape model behavior.

The course has two parts. The first part covers modern Python: packaging, strict typing with Pydantic, and asynchronous code for LLM calls. The second part covers prompt systems: how a model processes instructions, how to get structured output, and how to test and improve a prompt over time. Each part ends with a hands-on lab.

Requirements

• Working knowledge of Python
• Comfort running code from a terminal, with Git installed
• No prior experience with large language models or prompt engineering

What You'll Learn

• Package and structure a modern Python project for an AI system
• Validate LLM inputs and outputs with Pydantic and strict typing
• Write asynchronous Python code to call LLM APIs
• Explain how tokens, context windows, and inference parameters shape model output
• Design prompts with clear system and user instruction layers
• Enforce JSON-mode, schema-validated output from a model
• Apply few-shot prompting, prompt version control, and prompt compression
• Compare providers and API interaction modes for LLM systems
• Test and critique prompts, including A/B testing methods
• Describe the basics of agentic AI and foundation models

Hands-On

The course includes two graded labs. One lab covers modern Python for AI systems. The other lab covers prompt design. You build and test real code and real prompts.