Batch 1 (Jan 2027 Intake) — Starts Thu, 7 Jan 2027

What you'll learn

1. 1. Module 1: How Generative AI & LLMs WorkAI vs ML vs deep learning vs Generative AI · Tokens, context windows and next-token prediction · How models are trained: pre-training, fine-tuning, human feedback · Why models hallucinate and how to reduce it · Multimodal models: text, image, audio, video
2. 2. Module 2: The 2026 AI Model LandscapeChatGPT, Claude, Gemini: strengths and when to use each · Reasoning models vs fast models · Open-weight models (Llama, Gemma, Qwen, DeepSeek) and running AI locally (overview) · Choosing a tool: cost, privacy, language support, quality
3. 3. Module 3: Prompt Anatomy & Core PatternsThe 5-part prompt: role, context, task, format, constraints · Zero-shot vs few-shot prompting · Persona, audience and tone patterns · Format control: tables, bullet lists, word limits · Reusable prompt templates
4. 4. Module 4: Advanced Prompting TechniquesStep-by-step reasoning and when not to use it with reasoning models · Prompt chaining: breaking big tasks into steps · Self-critique and reflection prompts · Structured outputs: JSON and schemas for automation · System prompts and custom instructions
5. 5. Module 5: Testing & Improving PromptsWhy 'it worked once' is not enough · Building a 20-case test set in Google Sheets · Scoring with a rubric; A/B testing prompt versions · LLM-as-judge and its limits
6. 6. Module 6: Grounding & RAG (Intro)Why grounding matters: answering from your own documents · How RAG works: chunks, embeddings, retrieval (no-code explanation) · NotebookLM: sources, cited answers, audio overviews · Knowledge in custom GPTs, Gems and Claude Projects · Limits: stale data, bad chunks, privacy
7. 7. Module 7: AI Image GenerationHow image models work (diffusion and native image generation) in plain English · Image prompt formula: subject, style, composition, lighting, aspect ratio · Character and style consistency: references and editing · Text in images, posters and thumbnails · Commercial use, licences and watermarks
8. 8. Module 8: AI Video GenerationText-to-video vs image-to-video · Shot prompts: framing, motion, duration, sound · Tool tour: Veo, Sora, Kling, Runway (availability and cost) · Stitching clips into a short video
9. 9. Module 9: AI Voice & AudioText-to-speech options, including Indian languages · Pronunciation, pacing and emotion control · Voice cloning: consent, licences and misuse · AI music and sound effects: licences
10. 10. Module 10: No-Code AI Automation & AgentsWhat AI agents are: tools, memory, planning · Custom GPTs, Gems and Projects as mini-assistants · No-code workflows with Zapier / Make / n8n (overview demo) · MCP and connectors in plain English; agent safety
11. 11. Module 11: Responsible AI & Final ProjectCopyright, ownership and fair use of AI outputs · Privacy: India's DPDP Act basics, what not to upload · Deepfakes, labelling and platform disclosure rules · Bias and fact-checking checklist · Final project briefing and wrap-up

Final project demo: My GenAI Toolkit Project. Each learner delivers a mini campaign or explainer for a real topic: a tested prompt (with a 20-case evaluation sheet), a grounded NotebookLM brief, a consistent image set, a 15-30 second AI video with voiceover, and a completed responsible-AI checklist.