About the Course
Unit 1 — AI Web Apps: Build Your First Smart Interface
Summary
Students learn how to bring AI to life by building simple and interactive web apps. They explore Streamlit, connect to LLM APIs, and design clean user interfaces. This unit focuses on prompt engineering, helping students understand how to communicate effectively with AI models using system prompts, examples, constraints, and structured output.
Students will:
Build interactive AI-powered apps
Learn prompt engineering strategies
Understand system vs user prompts
Deploy a simple demo app
Unit 2 — LLM Agents: Teach AI to Think and Act
Summary
Students move from simple chatbots to rational LLM agents that can plan, decide, and take actions. They learn the difference between reflex rules and goal-oriented reasoning. Agents will read files, write files, make decisions, and produce JSON instructions the app follows—building the foundation of “AI that gets things done.”
Students will:
Build a rational agent loop (plan → act → reflect)
Design tools (functions) the agent can call
Handle invalid responses and add reliability
Create agents like homework planners or report generators
Sample Project: Homework planner
Unit 3 — Memory for AI: Make Your Apps Remember
In this unit, students explore how to give AI apps the ability to remember information from earlier interactions. Students learn the difference between short-term memory (keeping track of the current conversation or task) and simple saved memory (storing notes or preferences for later use). They design apps where the AI remembers what the user prefers, what tasks were done earlier, or what goals were set.
Students will:
Add short-term memory to their agents using conversation history
Build simple saved-memory features (e.g., notes, preferences, to-do lists)
Create apps that recall user information over time
Learn when AI should forget vs remember for safety and accuracy
Unit 4 — AI That Sees: Vision Apps with OpenCV & YOLO
Summary
Students learn how computers interpret the world through computer vision. They use OpenCV for transforming images and YOLO for detecting objects. Finally, they combine vision + LLM reasoning to build powerful real-world apps—from live object detectors to scene explainers.
Students will:
Capture and process images using OpenCV
Run YOLO object detection
Interpret results with an LLM
Build apps that understand the physical world
Example app: smart refrigerator to detect food and plan meals.
Your Instructor
Dr. Zhou
Location
2265 116th Ave NE, Bellevue, WA, USA
