Agentic AI Architecture

3.0 credits · Undergraduate

Agentic AI Architecture teaches students to build modern AI Agents and Applications, progressing from large language model (LLM) fundamentals to autonomous agents. Students first develop practical intuition for how LLMs behave, prompting, and reasoning. Then learn Context Engineering, like Retrieval-Augmented Generation (RAG), Agent Skills, and Agentic retrievals. The course then moves from AI that chats to AI that acts: students design agents with tools, memory, and reasoning, applying architectural patterns such as routing, orchestration, and reflection, and standardized interfaces like the Model Context Protocol. Hands-on work includes contemporary agentic coding tools (e.g., Claude Code and Codex). The course closes with AI strategy, governance, and ethics for responsible deployment. Programming projects are required throughout.