Inflearn - Lifetime Career Platform

This is a lecture directly from the author of <Prompt Engineering>. Master the fundamentals to skillfully utilize various AIs such as ChatGPT, Gemini, and Claude!

LLM: A pro quickly delivers core content, from essentials to practical tips!

The time when your imagination becomes reality—welcome to the world of 'Vibe Coding.' We prepared this course for those of you who may not know coding but know exactly what you want to create. With the emergence of AI tools like ChatGPT and Claude, people say that anyone can develop software. But what happens when you actually try? - "It writes simple code well, but once the number of files increases, the AI forgets what came before." - "It looks plausible, but when I run it, errors occur—and when I ask it to fix them, it gives me even more unrelated code." You’ve probably had experiences like these at least once. I experienced the same frustrations when I first started collaborating with AI. AI was an excellent 'assistant,' but it had limitations because it couldn’t see the entire forest. However, after discovering OpenAI Codex CLI and learning a new approach called 'Vibe Coding,' my development process completely changed. This course will solve the problems you’re experiencing while transforming you from someone who isn’t a coder into a director who commands AI agents. Equip yourself with the powerful tool Codex CLI, and join me on the exhilarating journey of Vibe Coding, where imagination becomes reality.

This is the edited live video from the LLM Application Experience Sharing Session with knowledge sharer Kang Byung-jin :)

Learn how to create a team of AI Agents with Grokbot that goes beyond simply chatting with AI to perform development, research, design, and operations tasks on your behalf. Build everything hands-on—from Agent design to Prompts/Rules, browser, terminal, and tool usage, Multi-Agent collaboration, Memory, Skills, Routines, cost optimization, and failure recovery—and ultimately complete a 24/7 AI Company system that performs repetitive tasks without requiring constant human instructions.

"How to Build a Smart AI Agent Inside Your Laptop for High-Performance Marketing" From brand analysis to creating ad videos, images, and card-style news content, all the way to data analysis—do it all at once! A marketer with 15 years of experience at a major corporation will provide you with the exact blueprints, skills, prompts, MCP integrations, and automation scripts of the 'marketing OS' they are actually using.

Create a program using the GPT API, and even generate images and videos!

Have you ever experienced any of these? You often hear the term "AI Agent," but when it comes down to it, you have no idea where to start. You tried giving ChatGPT a task, but it couldn’t access your company’s Outlook or Teams, so you ended up doing it manually again. You built a demo chatbot, but couldn’t take it to the level required for real-world work. You’ve heard of Power Automate, but don’t know what it has to do with AI agents. This course solves all four problems, in order. A true working agent, not just a chatbot that talks A chatbot provides answers; an agent gets the work done. In this course, you’ll use Microsoft Copilot Studio to build agents—without coding—that answer based on internal company documents, send emails, determine approval routes, and actually modify SharePoint data. Build five role-specific agents from start to finish Onboarding Check-In Bot (HR) — Proactively asks new hires about their readiness and answers based on internal company guide documents IT Help Desk Bot (IT) — Answers based on internal IT guides and forwards unresolved issues by email VOC Intake Bot (Customer Service) — Asks for all the necessary information based on the issue type and completes the intake Expense Approval Decision Bot (Accounting) — Checks item-specific limits, determines the approval route based on the amount, and forwards it by email Office Supplies Request Bot (General Affairs) — Checks SharePoint inventory, collects quantities through cards, and actually deducts the requested quantity from inventory Raise the level of control one step at a time This course does not simply list features. The entire course is designed around one central question: "How much should I delegate to AI, and how much should I control myself?" Lv1 Give instructions in plain language — Define rules through instructions Lv2 Provide reference materials — Ground answers in reference materials Lv3 Give it hands and feet — Use tools for external functions such as sending emails Lv4 Control the conversation — Design the conversation flow yourself using topics, variables, and conditions Lv5 Automate the work — Delegate decisions and processing to agent flows (Power Automate) Lv6 Change the data — Actually modify internal company data using SharePoint and adaptive cards What makes this course different Hands-on practice comes before explanations. In the second section, you’ll build your first agent without any explanation first, then take it apart step by step. We show failures first. You’ll deliberately encounter cases where a tool is not called and cases where the result is wrong despite no error, then find and fix the causes. The course was created only after running every exercise from start to finish on the actual interface. It is based on the Korean interface as of 2026, and also points out renamed menus (Tasks → Tools, Knowledge → Reference materials). We don’t stop at practice. You’ll publish the agents you build to Teams and M365 Copilot, and learn what to request from IT to connect them to an actual electronic approval system. Curriculum Introduction — Understanding AI Agents and Copilot Studio Agent Preview — Build one first; explanations can come later Instructions · Knowledge · Tools — IT Help Desk Bot Topics · Variables · Conditions — VOC Intake Bot Agent Flows (Power Automate) — Expense Approval Decision Bot Change the Data — SharePoint · Adaptive Card Office Supplies Request Bot Publishing and Operations — Teams publishing · Sharing · Monitoring Please check before enrolling You need a Microsoft 365 account provided by your company or school. Copilot Studio cannot be used with a personal account. The hands-on exercises for building and testing agents can be completed with the Copilot Studio free trial. Publishing to Teams requires an organizational license. Keep the structure you created today, change only the data and rules, and it becomes your own workflow.

Build a customer support agent from scratch that finds and answers from documents, selects tools to look up information, and actually creates approved tickets. Verify through real execution logs that each task is completed exactly once and through to the end, even if responses are interrupted or the same request arrives again.

Even when using the exact same AI, some people finish quickly while others struggle with it throughout the entire schedule. From providing AI service development consulting and lectures in corporate settings, I found that the difference ultimately comes down to whether they understand how LLMs work. Have explanations of LLMs always been so difficult that you couldn’t understand them? I explain the principles of LLMs and how to create a pre-training model in the simplest way possible. I’ll show you the principles that enable you to give AI precise instructions and verify its results.

This is a practical course on building a RAG-based document search system using the OpenAI API and Spring AI in a Spring Boot environment. You will learn the entire process of developing an AI assistant step by step, from the concepts of Prompt Engineering, Vector Stores, and Embeddings to rapid prototyping with Dify, MCP tool integration, multimodality, and guardrails. You can learn practical application strategies by comparing direct implementation with the use of no-code tools. 🆕 [Update] Section 4 now includes a supplementary lecture (1 hour) on TypeSafe’s Jev decision-making model, covering how to design an agent’s decision-making logic in greater depth.

From a single email to proposals and presentation materials—complete an entire work cycle with Claude Code.

The era of building things yourself is over; now is the era of delegation. With Codex’s simple instruct-check-refine loop, turn your ideas into real results.

You can learn about Text-to-SQL systems that generate and query SQL statements in real time in response to user questions.

You will learn about RAG, VectorDB, LangChain, OpenAI, sLLM, and fine-tuning, which encompass the most important AI-related technologies at this time.

This course is for non-developers who want to learn Claude Code but find the official documentation full of developer jargon, while YouTube focuses only on vibe coding (app/web development) and doesn’t explain how to use it for automating real-world work. Concepts are explained through comics and easy-to-understand analogies, with hands-on practice in a desktop app environment without using the terminal. It also walks you through three practical work-automation examples from start to finish: work automation, creating detailed product pages, and automating the production of card news, blog posts, and short-form videos.

[Source Code Provided] No coding knowledge required. Build it instantly in your browser using Google's latest tool (IDX) without any installation! Stop studying coding syntax! This is an ultra-fast, hands-on course where you'll build an RAG chatbot that perfectly understands internal company documents through AI conversation (Vibe Coding) in the Google IDX environment and deploy it live to the web.

This is an introductory course where you learn the core concepts of Claude Code through a 3D office analogy. From hiring a single robot in an empty office, giving it a brain and a heart, equipping it with tools, permissions, and memory, to eventually having multiple robots work together as a team—each concept unfolds as an office scene. You will master every concept by interacting with the digital course materials and complete your learning through hands-on practice on your own computer.

AI is only as good as you are. Learn vibe coding properly in three steps: how to use AI → essential foundational knowledge → project development. For those who want to learn vibe coding the right way, we deliver the core concepts that remain unchanged regardless of which AI you use, broken down into bite-sized pieces.

Document (SDD) → Code Automation. Make the system work instead of typing prompts. Build your own AI development agent with Skills, Sub-agents, and Commands, and complete monetization services through 5 real-world projects.