This is a live challenge that both individual customers and employer training customers
can participate in.
Individual customers are not eligible for a refund. Corporate training customers, please check the application and refund procedures below.
- Mandatory participation in both sessions
- Submit 2 of the 4 missions required
- Completion requires meeting both requirements 1 and 2
- If fewer than the required number, submit 3 out of 4 missions
- Proceed once the minimum number of 3 participants has been reached
- If fewer than 3 participants, carryover processing and refunds will be handled through consultation with the manager
- 2 screenshots in total (directly tied to the completion criteria)
- Once · Within 30 minutes of the start
- 2 sessions · Approximately 30 minutes before the end
- Camera must be ON when taking screenshots (with your face visible)
gov.partnership@inflab.com
Designing an AI Agent that can make decisions and take action independently
It doesn’t stop at calling an LLM.
You will build an Agent yourself that combines RAG and Tools to get the job done.
Building an AI Agent,
do you have
these concerns?
.
executes tasks by integrating LLMs, RAG, and Tools.
Six hours later, you’ll have an Agent that
gets your work done.
100% Live in real time
Build it live together with the knowledge sharer. Ask questions the moment you get stuck and solve them right there.
Close Mission Support
Each session has a mission. Pass with hands-on support from the operations manager and feedback from the knowledge sharer.
6 hours total, completed in 2 sessions
Two 3-hour sessions. Complete everything in one course, from Agent fundamentals to work automation Agents.
RAG System Development
Hands-on Project
Each class features
an agent for real-world tasks.
Workflow that
completes tasks from start to finish
Beyond LLM calls: Designing Agent workflows that complete tasks—from judgment and branching to completion—with a single instruction
When to use RAG and
when not to use it
Establish design criteria that distinguish situations requiring search from those where reasoning is sufficient, achieving both accuracy and speed.
Tool becomes the Agent's
action.
Complete an Agent that can be used immediately in practice with a ReAct structure that interprets results and determines the next action
Introduction to the Knowledge Sharer
CEO of Dapada Co., Ltd., an AI and data education company
Adjunct Professor, Department of Computer Software, Induk University
Numerous lectures and training sessions on LLMs/Agents for large corporations, financial institutions, and public institutions
Instructor for LLM courses on Inflearn
Book 『Introduction to LangChain: From RAG Chatbots to Agents』
Book: *Python Deep Learning with TensorFlow*
Book: 『Introduction to Python Deep Learning and Machine Learning』
A total of 4 modules
Apply them to your own work
Each module follows the structure: theory → practice → mission.
- LLM-based AI Agent Concepts and Principles
- LangChain Basics
- Core Prompt Engineering Techniques
- Build Your Own Expert Chatbot
- Understanding RAG Architecture
- Document preprocessing and Vector Database construction
- Developing a RAG-based Q&A Agent
- Building a company document search system
- Tool/Function Calling Concepts and Implementation
- Hands-on practice implementing various Tools
- ReAct Agent Design
- Create a Smart Assistant Agent
- Advanced Agents Built with LangGraph
- State Management · Memory Strategies
- Building a Workflow Automation Agent
A live course created
and operated by
Inflearn, where countless people learn
90% of the training fee reimbursed
If you are a small or medium-sized business, you can take this course with government support.
the rest is government-funded
Minimum guarantee for companies with fewer than 500 employees
settlement on your behalf.
Applications must be submitted in the company’s name by the person in charge of HR or training. This is for companies eligible for preferential support (employees currently enrolled in employment insurance), and applications may close early once the budget is exhausted.
Now delegate tasks to an Agent
Build an Agent structure that can make decisions and take action on its own.
Ask questions and resolve issues as soon as you get stuck.