Understanding the Fundamental Principles of Large Language Models (LLMs)

We explore how to create an AI judge by combining knowledge graphs, reasoning systems, and various agent frameworks such as CrewAI, LangChain, and AutoGen, which enable the implementation of LLM-based multi-agent technology.

Without a single formula or line of code, this book penetrates the essence of fundamental statistics needed for AI development and application.

By enabling junior developers to understand the entire development process from planning to code writing, deployment, and verification using artificial intelligence as a development tool, it helps them survive and thrive by acquiring the full-spectrum capabilities needed in the AI era.

This course systematically teaches prompt engineering, a core technology for effectively utilizing Large Language Models (LLMs) and generative artificial intelligence. It comprehensively covers everything from foundational theory to practical techniques, as well as the latest application cases and security/ethical issues, providing practical help to LLM-based service developers, data scientists, and AI planners alike.

Problematization is a term that can also be translated as problem-raising or problem-framing. It can also be translated as problem-setting or problem-definition. It is a concept that encompasses the process of questioning known facts such as requests or common sense from new perspectives, defining problems, and working out ways to solve these problems. While problematization should be the starting point of all development, it remains an insufficiently discussed topic in the development field. Carrying out projects or developing programs is, in fact, about creating plans to solve problems. In other words, it is related to problematization. To solve a problem, the problem must first be clearly defined. However, problems are usually given in the form of vague requests. Therefore, having the ability to transform vague requests into clear problems reduces unnecessary 'development waste,' facilitates smooth collaboration, and enables proper understanding of users' real needs. This course helps train 'problem-constructing thinking' through practical cases and tools.

What should be done when building a generative AI or LLM-based RAG (Retrieval-Augmented Generation) system, but the desired performance isn't achieved and there's no suitable solution? This lecture presents methods to improve RAG performance based on Cognitive Load theory. Through this lecture, you will understand the limitations of LLM context windows and learn how to effectively manage cognitive load in RAG systems. It is a practical-level theoretical lecture covering Chunk size and structure design, high-quality Chunk generation techniques, dynamic optimization, performance evaluation, and practical techniques.

You can learn the concepts and methods of 'service design' needed when planning applications or web services built on LLM, a type of generative artificial intelligence (generative AI).

"One name change, code transformed." "The key to readable code, 'naming for readability,' summarized from principles to examples." "Easy-to-read names begin collaboration."

This lecture is designed to be easily understood even if you have no prior knowledge of Secure Coding. After covering fundamental secure coding concepts, it focuses on web service security. This course was originally conducted as a special lecture hosted by OO University. It provides over 160 source codes with extensive comments, pinpointing core topics and techniques that can be applied immediately in the field.

You can quickly understand C language's basic concepts and grammar.

👨‍🍳 Like cooking code, neatly and deliciously! 『Clean Coding』 is a code cooking class for chef-like developers, where they joyfully learn clean coding through a cooking analogy. 🍽️

We introduce basic prompt patterns for coding and advanced API prompt patterns for leveraging artificial intelligence.

Starting from the origins of natural language processing technology, this provides a detailed explanation of the various language models developed in the journey leading up to the latest LLM models.

We will examine methods for collecting and analyzing user requirements to plan LLM services.

We explore the limitations of LLMs, methods to overcome them, and the latest research topics.

Let's explore the types of applications or web services that can be built using LLMs.

Explains why LLMs are needed, their technical background, and basic concepts.

Let's explore the types of applications or web services that can be built using LLMs.

We explore how to create an AI judge by combining knowledge graphs, reasoning systems, and various agent frameworks such as CrewAI, LangChain, and AutoGen, which enable the implementation of LLM-based multi-agent technology.

Explains why LLMs are needed, their technical background, and basic concepts.

This is a course covering the GitHub notebooks and bonus content from <Build a Large Language Model from Scratch> (Gilbut, 2025). GitHub: https://github.com/rickiepark/llm-from-scratch/ <Build a Large Language Model from Scratch> is the Korean translation of the bestseller <Build a Large Language Model (from Scratch)> (Manning, 2024) by Sebastian Raschka. This book provides a way to learn and utilize the operating principles of large language models by building a complete model starting from scratch with OpenAI's GPT-2 model.

You use ChatGPT, but haven't you found it difficult to explain why you get certain answers? "RAG, agents, fine-tuning... I know the terms but find it hard to explain them precisely" "I'm at a loss for words when I hear LLM-related terminology" "Explaining concepts in AI meetings is always vague" This course was created specifically for people like you. This course is a theoretical lecture designed to understand LLMs as a 'structure' rather than a 'tool'. It's not about how to use ChatGPT or Gemini, but about building a framework that allows you to explain why they work the way they do.

This is a lecture covering LLM theory and practical examples based on <Large Language Models, Just the Essentials!> (Insight, 2025).

The AWS Certified AI Practitioner practice exam course is a core summary program for passing the certification. First, you'll quickly review the fundamental concepts of AI/ML and generative AI, and learn about the features and use cases of AWS's major AI services that frequently appear on the exam (e.g., SageMaker, Rekognition, Polly, etc.). Based on this, you'll practice determining which AI solutions are appropriate for various real-world problem scenarios, and also cover content related to responsible use of AI models and data security. Finally, you'll learn problem-solving strategies by question type and prepare for the exam through practice mock tests. Through this course, you can efficiently prepare for the exam in a short amount of time.

Conquer everything about world-changing AI in just 4 lectures! Lecture 1. The True Identity of Artificial Intelligence AI is not magic. It's a digital brain created by analyzing human 'intelligence' like a mathematical formula! We thoroughly dissect the core structure of seemingly smart artificial intelligence. Lecture 2. Algorithm, AI's Creative (?) Thinking Method How does a pondering AI find answers? Experience firsthand the strategies of AI that plays Go, finds routes, and predicts your search results. Lecture 3. AI Hardware, The Engine That Turns Imagination Into Reality! The chip inside your smartphone is the heart of AI. We dissect everything from CPU, GPU, TPU, to even those unknown names. Lecture 4. The Great Clash of Machine Learning and Deep Learning! What on earth does it mean for AI to 'learn'? From classic rule-based AI to the currently hot ChatGPT! We trace the evolution of learning.

This course will teach you the concept of the Google Gemini model and how to use the Gemini API, and create various AI applications using Streamlit.

"Your first step to creating a customized LLM with LoRA-based lightweight fine-tuning!" This is an introductory hands-on course designed so that even those new to LLMs can easily follow along. We minimize complex theory and guide you step-by-step through the entire process: loading the model → applying data → training → comparing results. In a short time, you'll directly experience the workflow of cutting-edge lightweight fine-tuning techniques like LoRA and QLoRA, and gain an intuitive understanding of "how LLM fine-tuning works." Even without extensive resources, experience the satisfaction of creating an LLM specialized for your domain!

This course covers the entire process of evolving ChatGPT from a simple conversational AI to an intelligent AI Agent that performs actual work tasks. Using three core frameworks - LangChain, LangGraph, and LangSmith - you'll directly create AI Agents that can be immediately used in real-world applications.

Through this lecture, you'll be able to learn about the AI graduate school admission process and preparation methods from start to finish! I'll also share some great tips to increase your chances of getting accepted!

Create high-quality AI videos with Midjourney&Runway. Director Kim Dong-ah, who collaborated with Samsung, will personally guide you step-by-step through the AI video production process based on a virtual brand.

Time is constantly flowing, and data is constantly accumulating. In this dynamic world, time series data plays a vital role in every aspect of our lives. From the volatility of financial markets to subtle signals of climate change, time series data captures it all. Now learn how to interpret and leverage this powerful data using the power of Python and deep learning!

Learn how to utilize various local LLMs (Qwen, Gemma) and explore different techniques to efficiently evaluate and improve the performance of LLM systems.

This AI Literacy and Transformation Series (1) covers the following topics: 1. Data Center Cooling Innovation in the AI Era The Problem: The explosive power consumption of the latest AI chips (GB200, etc.) has pushed traditional air cooling methods to their limits, leading to AI performance degradation and massive operational cost issues. The Solution: Presents innovative liquid cooling technologies including Rear Door Heat Exchangers (RDCH), Direct-to-Chip (DTC) cooling, and immersion cooling. Additionally, provides strategies to simultaneously secure efficiency and sustainability through intelligent control systems that predict even leakage current and circular economy models that recycle waste heat for district heating. 2. The AI Paradox: Strengthening Critical Thinking to Awaken the Brain The Problem: The phenomenon of 'cognitive offloading'—outsourcing thinking to AI—is causing modern people's analytical thinking and literacy skills to rapidly deteriorate, creating a 'cognitive debt' problem. The Solution: Redefines AI not as a simple answer tool but as a tool for a 'cognitive gym'. Builds a system that makes you think first using critique, research, and synthesis agents, and strengthens metacognitive abilities as a 'helmsman' who determines direction rather than a passive passenger. 3. AI Memory Crisis: Next-Generation Architecture Design The Problem: The size of short-term memory storage (KV cache) required by AI is exploding, but expensive and limited-capacity HBM alone faces physical limitations in handling this data wave. The Solution: Presents a strategy for building a 'heterogeneous memory dream team' combining HBM, DRAM, CXL, etc. Provides solutions that dramatically lower corporate Total Cost of Ownership (TCO) by transferring smart system architecture design capabilities that balance capacity and cost rather than focusing on single component performance. 4. Palantir Ontology: Actionable Data Intelligence The Problem: Data is overflowing but fragmented, creating a 'data graveyard' problem where it fails to translate into actual decision-making and execution, blocking corporate innovation. The Solution: Gives data context and meaning through an 'Ontology' engine that maps physical assets to digital representations. Realizes data democratization where field personnel can directly command business without IT department assistance through 'Action Types (execution capabilities)' that immediately reflect analysis results into operational systems. 5. The Great Rewiring: AI Intelligence Stack Strategy The Problem: As the technology paradigm fundamentally shifts from the Internet era's 'deterministic computation' to the AI era's 'probabilistic prediction', existing business structures are being completely shaken. The Solution: Clarifies the reality of change through a '5-layer Intelligence Stack' map connecting semiconductors to applications. Provides a strategic decision-making framework for determining whether to Build or Buy AI capabilities, drawing a roadmap toward autonomous operating enterprises. 6. Forgetful AI: Maximizing Potential Through Feedback The Problem: AI's 'catastrophic forgetting' phenomenon where learning new information causes it to forget existing knowledge, and communication problems where unilateral commands prevent AI from performing properly. The Solution: Teaches 'coaching' and example-based dialogue methods instead of commanding AI. Transfers techniques for becoming a scientific training partner that enables AI to acquire new capabilities while maintaining existing knowledge through gradual feedback techniques that minimize behavioral change shock (KL divergence). 7. Engineering Financial Intelligence The Problem: Financial systems that remain stuck in past bookkeeping records and regulatory compliance cannot optimize capital allocation for an uncertain future. The Solution: Transforms into a 'dynamic scenario planning (war game)' system that computes thousands of variables in real-time using generative AI. Helps CFOs evolve beyond simple managers into strategic architects who design corporate value through 'Financial LLM' utilization methods that convert simple numbers into strategic signals.

In an era of explosive data growth and crucial real-time processing, edge computing has established itself as an essential technology, no longer an option. This course is designed to systematically teach everything, from the core concepts of edge computing to deployment strategies for its application in real business environments. Through this course, students will firmly solidify the fundamental principles of edge computing and acquire practical knowledge and know-how applicable to real-world tasks and projects. Take the most certain first step towards becoming an edge computing expert, a core technology in the future IT environment!

An introductory AI lecture tailored for those new to AI! Key concepts, from data to generative AI, are simply explained. The goal is practical understanding, enabling you to question and explain AI, not build it.

LLM and Agent, the fundamental technologies behind AI like ChatGPT. LLM is the technology that has come closest to AGI at this point in time. In this lecture, we explore LLM and Agent technologies from a general public perspective.

This course is a hands-on masterclass that completely conquers the evolution of Diffusion Models through papers and code implementation. You'll learn the core models of generative AI, including DDPM (Denoising Diffusion Probabilistic Model) and DDIM, by studying the paper principles and implementing them directly. We analyze step-by-step the background of each model's emergence, mathematical formulations, network architectures (U-Net, VAE, Transformer), training processes (Noise Schedule, Denoising Step), and the ideas that led to performance improvements. Students will directly code all models using PyTorch, gaining not just paper comprehension but 'practical skills to reproduce and apply' them in real-world scenarios. Additionally, by comparing the differences between models and their developmental flow, you'll clearly understand how they expand and evolve. This course integrates theory, code, and practice into one comprehensive journey, providing researchers, developers, and creators alike with a systematic way to master the evolution of generative models. Beyond simply 'reading' papers, start your experience of 'understanding and recreating' through direct implementation now.