LLMs, GitHub Copilot, prompt engineering, and multi-agent systems.
Tokens, transformer architecture, temperature, context window, and why LLMs generate text the way they do.
Using Copilot to generate PySpark, SQL, and Python code — effective prompting patterns and real examples.
Zero-shot, few-shot, chain-of-thought, and role prompting — with examples for code generation and data tasks.
Orchestrator, Planner, Executor pattern — how to design specialised agents that work together on a shared goal.