Phase 1
Phase 1: Foundations and Practical Modeling
M1-M4: why AI matters, how networks learn, how models train, and how vision models work.
Course Roadmap
AI Essentials
Concise Module Roadmap Summary
Roadmap At a Glance
The course moves from foundational AI fluency into language models, business workflows, and a final project that connects technical choices to practical value.
Phase 2
Phase 2: Language Model and Business Integration
M5-M8: how language models work, how to steer and ground them, and how to apply them in projects.
| Phase | Module | Focus | From the Figure | Why It Matters Next |
|---|---|---|---|---|
| Phase 1 | M1 | AI Development | Three waves of AI; jobs, society, human future | Frames why AI matters and why leaders need fluency. |
| Phase 1 | M2 | Neural Networks | Artificial neuron, gradient descent, core ML structures | Explains the basic learning machine. |
| Phase 1 | M3 | Model Training | Data preparation, model training, scaling law | Turns the machine into a practical workflow. |
| Phase 1 | M4 | CNNs | Convolutional layers, vision applications, human vs AI | Shows how the workflow adapts to visual data. |
| Phase 2 | M5 | LLM Foundations | Tokens, Word2Vec, embedding space, attention | Turns language into computable representations. |
| Phase 2 | M6 | LLM Architecture | NLP, RNNs, LSTM, Transformer | Explains the architecture behind modern LLMs. |
| Phase 2 | M7 | Generative AI and Agents | Prompt engineering, RAG, agentic AI, human-in-the-loop | Moves from model capability to business use. |
| Phase 2 | M8 | Final Project | Multi-agent AI collaboration for real-world application | Synthesizes the full roadmap in practice. |
Learning Frame
A Friendly Learning Frame
- You are not expected to become an AI engineer in one summer; you are building the literacy to ask better questions, make better judgments, and work creatively with AI.
- The technical pieces are connected: neurons lead to networks, networks lead to training, training leads to CNNs and LLMs, and LLMs lead to prompt/RAG/agent workflows.
- By the final project, the goal is not just to use AI. The goal is to understand enough to shape AI into something useful, trustworthy, and valuable for business.
In the end, we will gain a much deeper understanding of the three forces that drive AI productivity.
