CSCI 3347 AI Agent Engineering

This course focuses on engineering reliable AI agents and LLM-powered applications using modern software practices (Agile, DevOps, automated testing) adapted for LLM systems. Students implement the Analyze-Measure-Improve lifecycle, covering systematic error analysis, automated evaluation design (code-based and LLM-as-judge), core architectures (RAG, tool calling), UI/UX patterns, human-in-the-loop workflows, production observability, CI/CD pipelines, and various deployment strategies (TUIs, GUIs, client-server models). Emphasis is placed on building rigorously evaluated, production-ready AI agents. No prior machine learning background is required.

Credits

3

Prerequisite

Grade of "C" or better in CSCI 3340 or consent of instructor.

Schedule Type

Lecture

Grading Basis

Standard Letter (A-F)

Administrative Unit

Department of Computer Science