PART 1: Python for AI Development

Module 1: Python Fundamentals

• Python setup, core syntax, variables, data types and control flow
• Functions, and core data structures: lists, tuples, dictionaries, sets
• String handling and formatting for real-world text processing
• Lab: Build a command-line utility that reports word and character statistics

 

Module 2: Python Intermediate & Object-Oriented Programming

• Object-oriented programming: classes, objects, inheritance, encapsulation
• File handling and exception handling
• Regular expressions, iterators, generators and clean code practices
• Lab: Refactor a script into a class-based module with proper error handling

 

Module 3: Python for Data & APIs 

• NumPy and Pandas essentials for data manipulation
• Working with JSON and REST APIs, authentication and secrets management
• Git basics for collaborative development
• Lab: Consume a public REST API, transform the response with Pandas, export a dataset

 

Module 4: Advanced Python for AI Engineering 

• Decorators, context managers and asynchronous programming
• Type hints and Pydantic for structured data validation
• Debugging, logging and packaging for AI projects
• Lab: Build an async script that calls multiple APIs concurrently and validates responses

 

PART 2: Generative AI Development

Module 5: Foundations of Generative AI 

• The AI landscape, transformer architecture and attention mechanism
• How LLMs are trained: pre-training, instruction tuning, RLHF
• Core concepts: tokens, embeddings, context windows, temperature
• Model taxonomy: GPT, Claude, Gemini, Llama, Mistral and their trade-offs
• Lab: Call multiple LLM providers and a local open-source model, compare outputs

 

Module 6: Prompt Engineering Essentials 

• Anatomy of a production prompt, zero-shot, one-shot and few-shot prompting
• Structured output prompting: JSON mode, XML schemas, typed responses
• Prompt debugging and common mistakes
• Lab: Build a prompt template library that reliably returns structured JSON

 

Module 7: Structured Prompting & Autonomous Reasoning 

• Chain-of-thought, Tree-of-Thought and Graph-of-Thought reasoning
• ReAct prompting: the bridge into agentic AI
• Self-reflection, prompt chaining and meta-prompting
• Lab: Build a self-correcting reasoning pipeline

 

Module 8: GenAI API Integration & System Design Patterns 

• OpenAI, Anthropic and Gemini API deep dive, streaming and error handling
• Token budgeting, conversation memory and caching strategies
• Lab: Build a production-quality conversational API wrapper with streaming and memory

 

Module 9: Retrieval-Augmented Generation, Core 

• Why RAG, and full architecture: ingestion, chunking, embedding, retrieval, generation
• Chunking strategies, embedding models and vector databases (Chroma, Pinecone, Weaviate, FAISS)
• Semantic search mechanics and source attribution
• Lab: Build a document Q&A system that ingests PDFs and cites sources

 

Module 10: Advanced RAG & Multi-Modal Systems 

• Query transformation, re-ranking with cross-encoder models
• Hybrid search and multi-modal RAG across text, images, tables and code
• Metadata filtering and agentic RAG
• Lab: Upgrade the RAG system with hybrid search and re-ranking

 

Module 11: Fine-Tuning & Model Adaptation

• When to fine-tune: prompting vs. RAG vs. fine-tuning
• Parameter-efficient fine-tuning with LoRA and QLoRA
• Dataset preparation, training environment and quantisation for deployment
• Lab: Fine-tune a small open-source model with QLoRA and evaluate it

 

PART 3: Agentic AI Development

Module 12: Foundations of Agentic AI 

• What makes a system agentic: perception, memory, planning, action
• Agent types, and the reasoning-action loop: observe, think, act, reflect
• Lab: Build a simple reflex agent, then extend it into a planning agent

 

Module 13: Building AI Agents & Tool Use 

• Tool use with OpenAI function calling and Anthropic tool use
• ReAct and Plan-and-Execute agent patterns, handling failure modes
• Safe execution environments and agent evaluation
• Lab: Build a multi-tool agent that researches, executes code and produces a report

 

Module 14: Agentic Frameworks Deep-Dive 

• LangChain core and LangGraph for stateful, graph-based workflows
• CrewAI and AutoGen for multi-agent collaboration
• Model Context Protocol (MCP) and framework trade-offs
• Lab: Re-implement an agent pipeline using LangGraph and CrewAI

 

Module 15: Multi-Agent Collaboration & Orchestration 

• Multi-agent architecture patterns: hierarchical, sequential, collaborative, debate
• Agent memory systems and shared state management
• Handoffs, retries and the orchestrator-worker pattern
• Lab: Build a multi-agent research crew that searches, fact-checks and writes output

 

Module 16: Real-Time & Event-Driven Agents 

• Event-driven architecture: webhooks, queues, triggers, FastAPI listeners
• Streaming agent output over WebSockets
• Integrating agents with external APIs, calendars and databases
• Lab: Build an event-triggered productivity assistant

 

Module 17: AgentOps, Evaluation & Governance 

• Observability, tracing and evaluation metrics for agents
• Guardrails, safety filters and prompt/agent versioning
• AI governance essentials: EU AI Act, India’s DPDP Act, human-in-the-loop approval
• Lab: Set up a tracing and evaluation dashboard with human-in-the-loop approval

 

Module 18: Voice + Avatar Agents 

• Speech-to-text and text-to-speech pipelines for conversational agents
• Designing persona-driven voice or avatar agents
• Lab: Build a voice-driven agent that answers questions using a RAG backend

 

Module 19: Capstone Project 

• Teams choose a real-world track: finance, healthcare, retail, legal or open topic
• Guided build days with instructor mentorship
• Evaluation, security/risk assessment and presentation to a KnoaLabs instructor panel