Generative AI Training Course

What is Artificial Intelligence?
What is Generative AI?
Generative vs Discriminative Models
Neural Networks Introduction
RNNs, LSTMs, GRUs
Attention Mechanism
Transformer Architecture
Encoder, Decoder, and Encoder-Decoder Models
Autoencoders and Variational Autoencoders (VAEs)
Latent Space Representations
Generative Adversarial Networks (GANs)
Diffusion Models (Stable Diffusion concepts)
Evolution of Generative Models over Time

Introduction to NLP
Text Preprocessing and Tokenization
N-grams, Bag-of-Words, TF-IDF
Word Embeddings (Word2Vec, GloVe)
Introduction to BERT
Introduction to GPT Models
Sentence Embeddings and Document Embeddings
Hugging Face NLP Pipelines
Modern Tokenization: BPE, WordPiece, SentencePiece
NLP Evaluation Metrics

What is Prompt Engineering?
Components of an Effective Prompt
Zero-Shot, One-Shot, Few-Shot Prompting
Chain-of-Thought (CoT) Prompting
Role Prompting and Contextual Prompting
ReAct: Reasoning + Acting
Retrieval-Augmented Prompting
Self-Consistency Prompting
Tree of Thought (ToT) Prompting
Guardrails, Safety Prompts, and Fail-safe Design
Prompt Injection and Defense Strategies
Designing Prompts for Code, Reasoning, and Creativity

What is an LLM?
LLM Architecture
Attention and Multi-Head Attention
Token Embeddings, Positional Embeddings
LLM Parameters and Scaling Laws
Popular LLM Architectures: GPT, Claude, LLaMA, Gemini
LoRA, QLoRA, and PEFT Techniques
RLHF: Reinforcement Learning from Human Feedback
LLM Distillation
Evaluation of LLM Outputs
Using LLM APIs (OpenAI, Hugging Face, Gemini)
Understanding Context Length and Window Limitations

Concept of RAG in AI Systems
When and Why to Use RAG
RAG Architecture and Workflow
Embeddings for Retrieval
Vector Databases Overview
FAISS
ChromaDB
Qdrant
Pinecone
Document Splitting and Chunking
Query Engines and Retriever Types
RAG with LangChain
RAG with LangGraph
Multimodal RAG (text → image/video)
Evaluation of RAG Systems
Optimizing RAG for Latency and Accuracy

What is Agentic AI?
Traditional AI Pipelines vs Agentic Systems
Agent Architecture (planner, memory, executor)
Types of Memory (episodic, long-term, summarization)
Action Execution with External Tools
Multi-Agent Collaboration
Agent-to-Agent Communication Patterns
Model Context Protocol (MCP)
Popular Agent Frameworks
Designing Safe and Stable Agent Behavior

Introduction to CrewAI
Crew Roles and Task Distribution
Tool Usage in CrewAI
Creating Custom Tools
Memory in CrewAI (short-term, long-term)
Embeddings inside Crew Framework
Knowledge Systems in CrewAI
Planning, Reasoning, and Delegation
CrewAI CLI for Automation
CrewAI Flow for Workflow Design
Case Study: Fraud Detection Using CrewAI
Building Multi-Agent Teams for Real Use-Cases

Agentic RAG Overview
Agentic RAG with LlamaIndex
Introduction to n8n for Automation
Automating Workflows and Pipelines
Automated Email Classifier with n8n
Creating Business Automations with Agents
Deployment Techniques for GenAI Apps
Gradio
Streamlit
FastAPI
Monitoring, Logging, and Observability
Model Versioning and Rollbacks

Bias in Generative Models
Dataset Bias and Mitigation
Deepfakes and Synthetic Media
Prompt Injection Attacks
Jailbreak Attacks on LLMs
Safety Protocols for LLM Applications
Explainability and Transparency
Responsible AI Frameworks
Global AI Governance (EU AI Act, India AI Policies)
Ethical Model Deployment and Monitoring

Introduction to LangChain Framework
Prompt Templates, Chains, and Tools
Agents in LangChain
Retrieval with LangChain (RAG Pipelines)
Memory Types in LangChain
Working with LangChain Expression Language (LCEL)
Integrating Vector Databases (FAISS, Pinecone, Chroma)
Building Multi-Step Workflows
Deployment of LangChain Apps
Hands-On: Build a Full RAG Application

What is LangGraph and Why It Exists
Nodes, Edges, and State
Designing Graph-Based AI Agents
Persistence and Memory Handling
Multi-Agent Orchestration Using LangGraph
State Machine Design for LLM Agents
Error Handling and Guardrails in Workflows
LangGraph vs LangChain: When to Choose Which
Hands-On: Build a Stateful Multi-Agent System

Introduction to LangFlow UI
Visual Node-Based Workflow Creation
Building Prompts, Chains, and RAG Flows Visually
Custom Component Creation
Integration with Vector DBs & APIs
Rapid Prototyping of GenAI Apps
Deployment Options
Hands-On: Create a Visual RAG App

Core Concepts of LlamaIndex
Index Types (Vector, Keyword, List, Tree)
Storage Context and Service Context
Query Engines and Retrievers
Agentic RAG with LlamaIndex
Integrating Structured + Unstructured Data
Advanced RAG Techniques (Fusion, Re-Ranking)
LlamaIndex vs LangChain: Strengths & Weaknesses
Deployment Best Practices
Hands-On: Build Production-Grade RAG with LlamaIndex