inflearn logo

From the principles of generative AI to developing your own AI Agent - Multimodal, MCP, RAG, Agents SDK

Using ChatGPT is completely different from developing an actual generative AI service. This is because you must master everything from API calls to conversation state management, structured outputs, function calling, document-based RAG, MCP integration, and AI agent implementation. Based on extensive experience in data analysis and AI instruction, this course explains complex concepts through Python-centered practice. You will learn step-by-step—from the principles of generative AI to the OpenAI Responses API, multimodal features, File Search-based RAG, external tool integration, and the Agents SDK—to complete a fully functional AI application.

4 learners are taking this course

Level Basic

Course period Unlimited

Python
Python
Deep Learning(DL)
Deep Learning(DL)
AI
AI
ChatGPT
ChatGPT
prompt engineering
prompt engineering
Python
Python
Deep Learning(DL)
Deep Learning(DL)
AI
AI
ChatGPT
ChatGPT
prompt engineering
prompt engineering

What you will gain after the course

  • Developing Generative AI Applications Using the OpenAI Responses API

  • Implementing a Streaming Chatbot with Prompt Design and Conversation State

  • Implementing Stable AI Services Using Structured Outputs and Function Calling

  • Developing a multimodal AI app that combines image understanding and image generation

  • Building a Document-Based RAG System Using File Search and Vector Store

  • Implementing Tool-Calling AI Agents Using MCP and OpenAI Agents SDK

What will you learn in this course?

This course is a practical curriculum that completes Generative AI in three stages: Principle (Why) → API (How) → App Development (What).

Step 1 — Understand the principles of Generative AI through "intuition"

  • Fundamental difference between generative and discriminative models (distribution learning vs. decision boundary learning)

  • Transformer Architecture: Self-Attention, Positional Encoding, Multi-Head Attention

  • GPT's auto-regressive generation structure and Decoding Strategy (what temperature and top_p actually do)

  • The root causes of Hallucination and strategies for mitigation

Instead of simply saying "calling the API gives an answer," you need to understand why such an output is generated to confidently handle prompt design and parameter tuning.

Step 2 — Master the core features of the OpenAI API through hands-on practice

  • Text generation based on the latest Responses API, role design, and token management

  • Prompt Engineering: Principles and practical patterns of zero-shot / few-shot / Chain-of-Thought

  • Structured Outputs — How to receive LLM output as JSON for direct use in apps

  • Function Calling — How to make LLMs handle real-time data and external systems

  • Vision API & Image Generation (DALL-E / GPT-Image-1) — From the principles of ViT, CLIP, and Diffusion Models

  • Retrieval-Augmented Generation (RAG) using File Search and Vector Store

  • Conversation State Management and Streaming API

Step 3 — Completing Intelligent Agent Apps with the OpenAI Agents SDK

  • The evolution of LLM apps: from Chatbots to RAG Apps to Agentic AI

  • The core of the Agent, the ReAct loop (Think → Call Tool → Observe → Answer)

  • Handoffs (delegating tasks between agents), Guardrails (input/output safety measures), Sessions (conversation memory), Tracing (execution tracking)

  • MCP (Model Context Protocol) — The latest protocol for connecting agents to external tools and data in a standardized way

  • Final Project: Implementing Your Own Chatbot RAG Agent combining RAG + Tools + Handoffs

For every theoretical chapter, a paired Jupyter Notebook practice session (embeddings, autoregressive generation, Text Generation, image generation, Vision, Structured Outputs, Function Calling, Tools, Conversation State, Agent implementation) is provided via a GitHub repository, allowing you to run and modify the code yourself as you follow the lecture.

Here is what you will learn

Core Understanding of Generative AI and OpenAI API

Understand the principles of how Generative AI, LLMs, and Transformers work, and learn how to write effective prompts. Implement various AI functions such as text generation, image analysis, and image generation using the OpenAI Responses API.

Developing RAG and AI Agent Applications

Develop AI services connected to external data using Structured Outputs, Function Calling, Web Search, and File Search. Learn Vector Store-based RAG chatbots, MCP, and the Agents SDK, and complete a Streamlit-based AI agent.

Notes before taking the course

Practice Environment

  • Operating System: Windows or macOS recommended

  • Development Environment: Google Colab, Jupyter Notebook

  • Required Software: Python 3.10 or higher, latest web browser

  • Required Accounts: OpenAI account and API Key

  • Depending on your API usage, a small separate fee may be incurred.

  • A stable internet connection is required for downloading practice files and making API calls.

  • The lecture videos consist of screen recordings with Korean voice explanations. Since there are many code screens, we recommend watching on a PC or tablet with a large screen.

Learning Materials

  • We provide the lecture textbook PDF file and the Jupyter Notebook source code for practice.

  • OpenAI API example code is provided.

  • We provide sample data such as documents, text, and JSON for RAG practice.

  • We will complete the project by providing step-by-step explanations while directly executing and modifying the provided code.

Prerequisites and Important Notes

  • You can take this course even if you have no prior knowledge of generative AI or machine learning.

  • If you are familiar with basic Python syntax such as variables, functions, and lists, you will be able to proceed with the practice more smoothly.

  • The OpenAI API and its related libraries may have slight differences in screens or code depending on updates.

  • API keys should not be exposed externally or entered directly into the source code, but should be managed using a .env file.

  • Practice materials and source codes are for educational purposes only and unauthorized distribution is prohibited.

Recommended for
these people

Who is this course right for?

  • Developers and planners who have used ChatGPT but want to build their own apps directly using the API

  • Those who want to understand the principles behind "why" LLMs operate the way they do

  • Those who want to learn and experience the latest keywords such as RAG, Function Calling, MCP, and AI Agents through hands-on practice.

  • Working-level professionals considering the introduction of in-house AI services (chatbots, document search, task automation)

Need to know before starting?

  • Basic Python syntax (proficiency at the level of variables, functions, and loops is sufficient)

  • No prior knowledge of deep learning or mathematics required — necessary concepts are explained within the lecture.

  • Payment card registration is required to use the OpenAI API (pay-per-use, practice costs are around a few dollars).

Hello
This is YoungJea Oh

Inflearn Verified

Career Verified

4,968

Learners

475

Reviews

159

Answers

4.8

Rating

19

Courses

I am an AI specialist instructor teaching Artificial Intelligence and Python based on over 30 years of IT field experience. I develop and teach practical-oriented curricula, including deep learning, NLP, LLM fine-tuning, LangChain/LangGraph-based AI agents, and AI-driven development (AIDD). I translate more than 30 years of development and operations experience gained at Hyundai Engineering & Construction's IT department, Samsung SDS, and Citibank Korea into realistic, hands-on lectures. Currently, I am conducting AI courses at institutions such as KOSA, KOSTA, and KITRI.

Homepage Address: https://ironmanciti.github.io/

More

Curriculum

All

40 lectures ∙ (7hr 3min)

Course Materials:

Lecture resources
Published: 
Last updated: 

Reviews

Not enough reviews.
Please write a valuable review that helps everyone!

YoungJea Oh's other courses

Check out other courses by the instructor!

Similar courses

Explore other courses in the same field!

Limited time deal

$27,500.00

50%

$42.90