Full Curriculum
The Generative AI & AI Application Development Curriculum
14 sessions, 42 hours, interleaved so you build within 48 hours of learning each concept. Learning philosophy: Experience → Understand → Build → Ship → Show.
This is the exact, session-by-session schedule taught in the Generative AI course. Prefer the short version first? See the course overview or jump straight to the 3 hands-on projects.
Pre-course
Session 0 — Launch & Setup
30 min, online- Install Python + VS Code, verify with a "hello world" script
- Create free API keys (Groq / Gemini free tier)
- Join the course group where every "ship it" screenshot gets posted
- Receive the course roadmap with outcome-based module names
Session 1
"Talk to a Machine": Your First AI App + AI Fundamentals
3 hrs- Build a working chatbot in an API playground in the first 20 minutes — before any theory
- AI vs ML vs Deep Learning vs Generative AI
- What is an LLM? Training vs inference, tokens & tokenization, context windows
- Activity: classify 10 everyday products as AI / ML / GenAI
Session 2
Inside the Machine: How LLMs Actually Work
3 hrs- Foundation models vs applications; a tour of GPT, Claude, Gemini, Llama, and Mistral
- How LLMs generate text — next-token prediction, live demo
- Transformer architecture, intuition only, zero heavy math
- Temperature & generation controls — same prompt at temperature 0 vs 1
- "Break the AI" game: make a model hallucinate, then discuss why
Session 3
LLM Wrap-up + The Art of Asking: Prompt Engineering I
3 hrs- Hallucinations, model limitations, multimodal AI overview
- Zero-shot, one-shot, and few-shot prompting
- Role & context prompting, instruction hierarchy, structured JSON outputs
- Build a resume analyzer prompt and an email generator prompt
Session 4
Prompt Battle + Your AI Toolbox
3 hrs- Prompt templates, decomposition, evaluation & iteration; reducing hallucinations through prompting
- Prompt Battle: a live class competition, judged on identical test cases — the leaderboard begins here
- ChatGPT, Gemini, and Claude: strengths and when to use which
- AI coding assistants, AI-powered IDE workflows, Perplexity and NotebookLM for research
Session 5
Tools Wrap-up + First Code: Calling LLMs from Python
3 hrs- Full AI-assisted dev workflow demo: requirement → design → code → test → debug → document
- Plain Python scripts first — no framework yet; system vs user messages
- Model selection, temperature settings, token usage & cost awareness
- Ship it: a Python script that answers a question via an LLM API
Session 6
Building a Real Backend + Project 1 Kickoff
3 hrs- FastAPI-based AI backend from provided boilerplate, streaming responses
- Rate limits, error handling & retries — the same patterns a production API needs
- 🚀 Project 1 begins: AI Study Assistant — question answering, concept explanation, note summarization
Session 7
Project 1 Finish + Why AI Needs Your Documents
3 hrs- Finish the AI Study Assistant: MCQ generation, study-plan generation; ship a demo screenshot
- Why RAG is needed — Retrieval-Augmented Generation concept introduced
- Embeddings, explained as GPS coordinates for meaning
Session 8
Build a ChatGPT That Knows Your Documents: RAG Deep Dive
3 hrs- One stack end to end: Chroma + one embedding model + one LLM — with LangChain, LlamaIndex, FAISS, and pgvector covered briefly as "alternatives that exist"
- Document ingestion & chunking, metadata & semantic search
- Retrieval & context injection — the full RAG pipeline; RAG vs fine-tuning
- Live demo: the instructor's full pipeline on a real PDF, then students replicate it
Session 9
🚀 Project 2: Chat With Your Documents
3 hrs- PDF upload & parsing, chunking + embeddings, vector search with Chroma
- RAG-based answers over your own textbooks, notes, or papers
- Leaderboard event: first working end-to-end RAG pipeline wins points
Session 10
Project 2 Finish + Giving AI Hands: Agents & Tools
3 hrs- Finish Project 2: relevant source/reference display; ship a demo video
- LLM vs AI agent — an agent is an LLM with hands, not just a mouth
- Tool/function calling, agent loops & tool selection
Session 11
Agents Finale + AI That Sees and Speaks
3 hrs- Planning & multi-step tasks, memory concepts, external APIs as tools, agent limitations
- Mini-build: a calculator + search-tool agent
- Multimodal AI: image generation, vision models, speech-to-text, text-to-speech, OCR
- Live demo chain: voice → text → LLM → voice
Session 12
Making It Real: Production AI + Capstone Kickoff
3 hrs- Architecture (backend + LLM + data layer), caching, rate limiting, authentication
- Logging, monitoring, prompt versioning, latency & cost management
- Prompt injection challenge: attack a demo app live, then learn the defences
- 🚀 Capstone begins: AI Career Assistant — students upload their own resume on day one
Session 13
Your AI Career + Capstone Build
3 hrs- AI Engineer, GenAI Engineer, AI Product Engineer, ML Engineer — roles compared
- Live job-posting walkthrough: real listings and salary ranges mapped to what you've built
- Portfolio building: turning course projects into problem → architecture → tech → outcome case studies
- Capstone build: resume analysis, job-description analysis, skill-gap detection
Session 14
Capstone Sprint + Demo Day
3 hrs- Personalized learning-plan generation, mock interview + answer evaluation
- Tool calling + RAG integration + external APIs, combined in one capstone
- 🎤 Demo Day: every student presents their AI Career Assistant to peers, faculty, and an invited industry guest
- Certificate awarded + a personalised "what to build next" roadmap
Hour accounting
How the 42 hours break down
| Module | Hours | Where |
|---|---|---|
| AI Fundamentals | 3 | Session 1 |
| GenAI Fundamentals | 4 | Sessions 2–3 |
| Prompt Engineering | 3 | Sessions 3–4 |
| AI Tools & Productivity | 3 | Sessions 4–5 |
| LLM App Development | 4 | Sessions 5–6 |
| RAG & Knowledge Systems | 4 | Sessions 7–8 |
| AI Agents & Tool Calling | 3 | Sessions 10–11 |
| Multimodal AI | 2 | Session 11 |
| Production AI | 2 | Session 12 |
| AI Career & Future | 2 | Session 13 |
| Project 1 — Study Assistant | 3 | Sessions 6–7 |
| Project 2 — Chat With Docs | 4 | Sessions 9–10 |
| Capstone — Career Assistant | 5 | Sessions 12–14 |
| Total | 42 |