Hands-on Generative AI workshops — from fundamentals to autonomous agents. Tokenization, prompt engineering, image generation, RAG, AI agents, bias detection, and safety guardrails. Fully self-paced. Runs anywhere: cloud, local GPU, or fully offline.
By completing this workshop series, you will be able to:
- ✅ Explain how LLMs, Transformers, and diffusion models work
- ✅ Tokenize text and understand how AI reads language
- ✅ Generate text and images using open-source AI models
- ✅ Apply prompt engineering techniques for better AI output
- ✅ Build RAG pipelines with vector embeddings and semantic retrieval
- ✅ Build AI agents with tool-calling (ReAct pattern)
- ✅ Run local LLMs with Ollama (zero cloud dependency)
- ✅ Detect and mitigate bias in ML models using AIF360
- ✅ Implement safety guardrails on AI outputs
| # | Workshop | What You'll Build | Format | Time | Recording |
|---|---|---|---|---|---|
| 1 | GenAI Fundamentals | Tokenizer, text generator, prompt lab, local-vs-cloud comparison | Colab Notebook | ~90 min | |
| 2 | Building with GenAI | Image generator, RAG pipeline, mini AI agent with tools | Colab Notebook | ~2 hrs | |
| 3 | Healthcare AI Agent | Autonomous healthcare agent with vision, RAG, bias detection, safety | Python Scripts | ~2 hrs |
Each workshop is self-contained. Clone the repo and pick any workshop:
git clone https://github.com/bhaskarjha-com/genai-workshops.git
cd genai-workshops
# Workshop 1 or 2: Open notebook in Google Colab or Jupyter
cd 01_genai_fundamentals
# Open Workshop_1_GenAI_Fundamentals.ipynb in Colab
# Workshop 3: Run Python scripts directly
cd 03_healthcare_ai_agent
pip install -r requirements.txt
python 01_multimodal_medical_ai.pyNo GPU? No API key? Workshops 1 & 2 run on free Colab GPUs. Workshop 3 has a Demo Mode that works fully offline with pre-recorded outputs.
┌────────────────────────┐ ┌────────────────────────┐ ┌──────────────────────────┐
│ 01 FUNDAMENTALS │ │ 02 BUILDING │ │ 03 HEALTHCARE AGENT │
│ │ │ │ │ │
│ • Transformers │ │ • Image generation │ │ • Multimodal vision │
│ • Tokenization ★ │───▶│ • RAG pipelines ★ │───▶│ • Autonomous agents ⭐ │
│ • Text generation │ │ • Mini AI agent ★ │ │ • Bias & fairness │
│ • Prompt engineering │ │ • Local vs cloud │ │ • Safety guardrails │
│ • Local vs cloud │ │ │ │ │
└────────────────────────┘ └────────────────────────┘ └──────────────────────────┘
Foundations Applications Production-Grade
Recommended order: 1 → 2 → 3 (each builds on the previous). Can I skip? Yes — each workshop has its own README with prerequisites.
| Requirement | Workshop 1 | Workshop 2 | Workshop 3 |
|---|---|---|---|
| Python 3.10+ | ✅ (via Colab) | ✅ (via Colab) | ✅ (local) |
| Google Account | ✅ | ✅ | Optional |
| GPU | ❌ Free Colab GPU | ❌ Free Colab GPU | Optional (demo mode) |
| Ollama | ❌ | Optional | Optional |
| API Key | ❌ | ❌ | Optional (free Gemini) |
| Prior AI knowledge | ❌ None | Workshop 1 | Workshops 1 & 2 |
Each workshop includes a guides/ folder with ready-to-use teaching materials:
| File | What It Contains |
|---|---|
INSTRUCTOR_GUIDE.md |
Timeline, teaching hooks, Q&A prep, emergency procedures |
PRESENTATION_SCRIPT.md |
Minute-by-minute delivery script (Workshop 3) |
RESOURCES.md / STUDENT_RESOURCES.md |
Curated post-workshop learning paths |
Delivery modes:
- 🎓 Live workshop: Follow the instructor guide, share screen, run code together
- 📖 Self-paced: Students follow the README and run cells/scripts independently
- 📋 Demo mode (W3): Pre-recorded outputs, works without any API keys or GPU
genai-workshops/
├── README.md ← You are here
├── LICENSE (MIT)
├── CONTRIBUTING.md
├── CHANGELOG.md
├── .gitignore
│
├── 01_genai_fundamentals/
│ ├── README.md Student guide
│ ├── requirements.txt
│ ├── Workshop_1_GenAI_Fundamentals.ipynb
│ └── guides/
│ ├── INSTRUCTOR_GUIDE.md
│ └── RESOURCES.md
│
├── 02_building_with_genai/
│ ├── README.md Student guide
│ ├── requirements.txt
│ ├── Workshop_2_Advanced_GenAI.ipynb
│ ├── Running_Local_Models_Deep_Dive.ipynb
│ ├── Agent_Programming_Guide.ipynb
│ └── guides/
│ ├── INSTRUCTOR_GUIDE.md
│ └── RESOURCES.md
│
└── 03_healthcare_ai_agent/
├── README.md Student guide
├── requirements.txt
├── .env.example
├── workshop_config.py
├── 01_multimodal_medical_ai.py
├── 02_medical_rag_frontier.py
├── 03_healthcare_agent_frontier.py
├── 04_bias_and_fairness.py
├── 05_safety_and_guardrails.py
├── sample_chest_xray.png
└── guides/
├── INSTRUCTOR_GUIDE.md
├── PRESENTATION_SCRIPT.md
└── STUDENT_RESOURCES.md
This is educational material. Healthcare scenarios in Workshop 3 use simulated data for teaching purposes. Not a medical device. Not for clinical use.
Found a bug? Have a suggestion? See CONTRIBUTING.md.
MIT License — free to use, modify, and distribute.
Created by Bhaskar Jha for the VIT University (BMESI & BMSA × IIIC) workshop series on Generative AI, 2026.
Built with: Hugging Face · Google Gemini · Ollama · ChromaDB · IBM AIF360