Full Stack Generative and Agentic AI with Python
- Source URL: https://www.udemy.com/course/full-stack-ai-with-python/
- Instructors: Hitesh Choudhary, Piyush Garg
- Type: Online Course (Udemy)
- Tags: AI Python AgenticAI RAG MCP DevOps
Overview
A comprehensive bootcamp for building modern AI applications, ranging from foundational Python to advanced AI engineering including Agents, LangGraph, and RAG pipelines.
Key Modules
1. Foundational Skills
- Python: Asyncio, threading, multiprocessing, Pydantic, and OOP.
- DevOps: Docker containerization, Git/GitHub workflows.
2. AI Fundamentals
- LLMs: Attention mechanisms, transformers, tokenization, embeddings.
- Prompt Engineering: Zero-shot, few-shot, CoT, structured prompting.
- Integration: OpenAI, Gemini, local models via Ollama/Hugging Face.
3. Retrieval-Augmented Generation (RAG)
- LangChain: Document loading, splitting, and retrieval.
- Scaling: Redis/Valkey queues, asynchronous processing with FastAPI.
4. AI Agents & Graphs
- Agentic Workflows: Building agents from scratch.
- LangGraph: Stateful AI systems using nodes and edges.
- Memory: Short-term, long-term, episodic, and semantic memory using Mem0, Vector DBs, and Neo4j.
5. Ecosystem & MCP
- Model Context Protocol (MCP): Building and implementing MCP servers to enhance tool-calling.
Lecture Logs
Lecture: Experience in Writing Python Code
- Topic: Basic Python syntax and environment setup.
- Key Takeaways:
- VS Code setup with Chai Theme and Pylance.
- Functions as “boxes” for instructions.
- Python’s reliance on 4-space indentation for structure.
- Logic-first approach: “If you can read English, you can read Python.”
- Reference: Python-Basics