ChatGPT
New chat
Pinned
Proposal draft exercise
Automate meeting notes
Build an internal FAQ GPT
24 learners · practice accounts
ChatGPT for work
From work prompts to building GPTs
SMB Channel PartnerGenerative AI training for companies and public institutions
AI training that changes how work gets done
We assess your organization's AI level first, design around real tasks by role, and run 70% hands-on. One team stays with you until it sticks in daily work.
Talk to usCompanies and institutions
learn AI with Imagination Group
AI training programs
Learn on your own work,
use it from day one
Imagination Group's AI training programs are hands-on courses that get employees in companies, public institutions and factories using generative AI such as ChatGPT, Claude and Gemini in their own jobs. We assess the organization first, then design by level, from AI literacy for everyone to role-specific courses and advanced AI agent and automation tracks. Seventy percent of each course is practice, and we support adoption after the class.
One tool, in depth
If your organization has already chosen its tool, a 16-hour intensive course is the right fit.
Our principles
Training that changes how work gets done
The goal is not a certificate but daily use at work
From AI literacy talks to the AI Basic Act, ethics and role-based prompt practice, every course follows the same principles
No extra studying needed —
anyone becomes an AI practitioner
Facing so manyroleslevelsformats, hesitation disappears
Assessing your organization…
Literacy track
- Planning 16hChecking
- Front desk 8hChecking
- Executives 4hChecking
We assess your organization first
and show only the courses that fit each role
Get a free AI assessment
Custom curriculum
Practice with your own documents
30
30 days after
Adoption program
Not slide decks —
courses built on your own work documents
Browse the programs
Checking institutional requirements…
Training operations checklist
- AI usage rules · security
- Paid practice accounts
- Backup devices · lines
- Three-stage instructor vetting
- Final report
Security, accounts and devices checked before day one,
so practice starts in the first hour
See operating requirements
01Assess
We look at your level first,
so no one gets the same course by default
Before training we assess each employee's AI usage level and the tasks of their role. From AI literacy for everyone to AI ethics, the AI Basic Act and executive AX programs, courses are placed by four levels and by role and industry.
Recommended placement
Courses, hours and audiences placed from the assessment
- Planning · policyChatGPT Master · advanced LLM16h
- Front desk · serviceChatGPT prompt basics8h
- ExecutivesExecutive AI strategy workshop4h
Place by role
From the assessment we place the right course and hours for each role. No one gets the same course by default, so practice connects straight to real work.
AI level assessment
Four levels · by roleStrategy and governanceAdvanced and productionApplied to the jobAI literacy
Split by level
Employees are grouped into four AI levels so you can see your organization's spread — where to start, and who needs an advanced track.
Every organization uses different tools,
so training should fit too Programs
Claude Projects
MMinji KimPlanning
SSeojun ParkSales
HHaneul LeeHR
Project knowledge
Artifact preview
Claude for work
Automate document work with Projects & Artifacts
Gemini
Summarize this doc in one paragraph
Draft a scheduling email
Explain this spreadsheet formula
Gemini에게 물어보기
Gemini for Google Workspace
Practice inside the docs, mail and sheets you use
ai-agent-lab
$ ai-agent-lab run
model: gpt · claude · gemini
task: internal document search agent
tools: function calling · RAG
env: lab environment provided
✓ ready
$
For developers and data teams
API & agent hands-on course
Training ops checklist4/5
- Instructor credentials
- Security pledge · air-gapped lab
- Contract · proposal documents
- Final report · satisfaction
- Attendance · certificates
For public institutions
Operating paperwork prepared with you
RosterCSV
- Computer Science · Year 2
- Business · Year 3
- Staff · Registrar
- Staff · Research support
For universities and schools
Practice accounts and tracks by department
02Hours
How many hours you need,
shown up front
Hours are set by the share of practice and organization-level work, not by difficulty. From a 2-hour AI talk to a 40-hour company-wide program, classroom and online AI training can be mixed.
Concepts and strategic judgmentHands-on work and building the system
2 hours
DeliverableLecture
Our AI usage guide
1Safe to use AI for
2Never paste in
33-step output check
Everyone
AI fundamentals and literacy
4 hours
DeliverableDiscussion
AI adoption roadmap
Q1Q2Q3Q4
Pilot team
Company rollout
Impact review
Executives
Executive AI strategy workshop
8 hours
DeliverableHands-on
Service reply prompt
[Role][Context][Ask][Format]
Front-line service
ChatGPT prompting for beginners
16 hours
DeliverableHands-on
Policy review GPT
Summarize the key issues in this bill
Cited clausesSimilar cases
Planning and policy
ChatGPT mastery and advanced LLM use
16 hours
DeliverableHands-on
Auto budget tracker
TeamSpentLeft
78%
54%
91%
36%
Admin and general affairs
Workflow automation with AI and Excel
24 hours
Deliverable
Poster
Card news
Short video
ThumbnailProduction
Campaign asset set
PosterCard newsShort videoThumbnailMarketing and comms
Image and video generation with AI
40 hours
DeliverableProject
AI security checklist
- Mask personal data
- Approve external sharing
- Separate access
- Keep usage logs
IT and information security
Company-wide AI program with security practice
40 hours
DeliverableBuild
Knowledge base structure
RulesManualsReports
Knowledge baseOrganization-wide data
Restructuring the organization's data
Six delivery formats
Mixed to fit your schedule and goals. We are based in Seoul and have delivered training for institutions outside it, including Daejeon and Jeju. Ask us about on-site delivery in other regions such as Busan, Daegu, Ulsan, and Gwangju.
Two to four hours for all staff or executives. An introduction to where generative AI is going and how your industry is using it.
AI talks and seminars
Eight to sixteen hours. Participants bring their own work documents and finish role-based tasks end to end.
In-person hands-on training
Live online coaching and custom video courses for teams that cannot gather in one room, such as branch offices and shift workers.
Online AI training
A short, intensive program that brings new hires and key staff up to speed, ending with a team project and presentation.
AI bootcamp
In-house or institution-wide contests where people apply what they learned to real problems. We run them from planning through judging.
AI hackathons and competitions
When your own program needs an AI or ChatGPT instructor, we place one who has passed our three-filter check.
Instructor placement
03Adoption
The course ends not on the last class day
but on the day it sticks at work
A learning report shows before, during and after the course on one screen, and a 30-day adoption program confirms that what was learned becomes real work.
AI training report · Cohort 2
Learners
48
Assignments
46
Using after
41
Daily submissions
NameDept.BeforeAfter
MMinji KimPlanningLv.2Lv.3
SSeojun ParkSalesLv.1Lv.2
HHaneul LeeHRLv.2Lv.3
YYujin ChoiAdminLv.1Lv.3
DDoyun JungPRLv.2Lv.3
✓Assignment feedback✓Q&A✓Usage report
Learning Report
From assessment scores to assignments and usage,
before, during and after the course
on one screen.
After 30 Days
Not the last class day,
but the day it sticks at work
is when the course ends.
About the adoption program04Institutional requirements
Will it run inside your institution?
Five answers
Practice is essential, but institutional data must stay safe. Security, accounts, devices, instructors and reporting are prepared together before the first session.
AI usage rules · security guide
Input rules for personal and internal data and de-identification steps aligned with your security policy, plus enterprise security settings reviewed with your staff.
Paid practice accounts · short-term licenses
Team accounts subscribed only for the training period, issued before day one, run separately so institutional data never stays in personal accounts, and retrieved afterwards.
Backup devices · backup lines
Our own practice laptops, wired and wireless backup from three carriers and portable satellite internet for instant failover. Access blocks are checked before the first session.
Instructor vetting documents
Instructors vetted on real-world builds, public-sector teaching and 1:1 coaching, submitted in the format your institution requires.
Final report · usage report
Attendance, assignments, level change and post-training usage per cohort, compiled in your institution's reporting format.
The three filters
Teaching experience alone
doesn't get an instructor in
Teaching history isn't enough. Only people who clear all three — real implementation, public sector teaching, and one-to-one coaching — get assigned.
- 01Real implementation experienceWe check for automation that actually runs, not a teaching record. Only people who designed the workflow and connected the tools themselves get through.
- 02Public sector and corporate teachingWe check for experience running training for public sector staff — cohort delivery, security compliance, and reporting.
- 03One-to-one feedback and setupWe check whether they can give per-learner feedback on assignments and support installs and account setup, and whether they have run Q&A sessions alone through a multi-session course.
Each criterion is judged against a CV, delivery record, and work samples, and assignments are confirmed with the client before the course starts.
Vetting criterion
Real implementation experience
Vetting criterion
Public sector and corporate teaching
From assessment to adoption,
four steps are enough
One team is accountable from consultation to delivery.
- STEP 1
Assess
We measure your AI capability and set the target together.
- STEP 2
Design
We build the course around the work your teams actually do.
- STEP 3
Train
Hands-on sessions turn the tools into working habits.
- STEP 4
Apply
A 30-day adoption program carries it into the day job.
See all operating requirements
Using AI safely
Internal documents, personal data, de-identification, and institutional AI security
Key considerations
- No standard for deciding whether internal material can go into an AI
- No procedure for de-identifying personal or sensitive data
- Concern about input data being used for training or sent overseas
- The institution's security policy isn't connected to any rules on AI use
What we prepare
- AI ethics and risk guidelines — We review privacy, copyright, and bias risk, then write usage rules for your institution
- Hands-on Enterprise security setup — We walk through the OpenAI and Claude Enterprise security settings with your lead
- Input masking and auditing — AI Switch DLP masking, conversation audits, and IP allowlisting
- AI security management dashboard — Patent pending for AI security management in the public and corporate sector (10-2026-0033627)
A practice environment that works
Paid AI accounts, short-term licenses, and accounts and software set up in advance
Key considerations
- The free tier limits how far the practice can go
- Accounts are only needed for the course, but approval runs on annual, overseas payments
- Practising on personal accounts leaves institutional material in them
- Creating accounts and installing software eats into the first session
What we prepare
- Paid accounts for the course — ChatGPT, Claude, Perplexity and 25+ others, at a fixed price in KRW
- Short-term licenses — Subscribed only for as long as the course runs, with a per-account delivery note (PDF)
- Separate accounts for the course — Issued separately so institutional material never lands in a personal account
- Accounts and software set up in advance — Accounts created and software installed before day one, so practice starts in the first hour
Three layers of backup for the practice room
So a failure doesn't stop the class, we put a fallback in place for each point of failure before the course opens.
Hardware
PC failure or not enough seats
- We own the practice laptops — brought in the moment a lab PC fails or seats run short
- A second set of hardware set up alongside the room, sized against the seat count before day one
Accounts and software
Sign-in errors or a missed install
- AI tools installed and configured in advance — paid team accounts issued in bulk before the course (25+ tools)
- A bulk sign-in check the day before each session, with spare accounts ready to reissue on the spot
Network
Outage or slow connection
- Wired and wireless backup across all three carriers, plus mobile satellite internet, switched over the moment a line drops
- A connectivity check before the course opens, confirming ChatGPT, Claude, and Gemini aren't blocked
Hardware, accounts, and network requirements are confirmed against a checklist with the institution before the course is finalized.
Audiences
Choose the training for your audience
AI training designed in the language and data of companies, public institutions and factories. From new hires, managers and executives to civil servants.
For Company
Corporate AI training
Hands-on training that changes how employees work. Role-based scenarios from ChatGPT through AI agents and advanced LLM use.
See programs
For Public
Public sector AI training
AI training for civil servants and public institutions. Official document automation, AI literacy, and on-site expert placement — with security handled from the start.
See programs
For Manufacturing
Manufacturing AI training
Training that solves problems with shop-floor data. Smart factory basics through predictive maintenance, vision inspection, process optimization, and physical AI.
- Smart factory basics
- Predictive maintenance
- Vision inspection and defect detection
- Process optimization
- Physical AI
FAQ
Frequently asked questions
What people ask most before a consultation.
How is the AI training program structured?
It runs in four stages: assess, design, practice, adopt. We assess how the organization uses AI today, design courses around its industry and roles, teach with seventy percent hands-on practice, and support adoption afterward. Courses span four levels, from AI literacy to role-based use, and on to AI agents and advanced LLM work.
Can the training run online?
Yes. We run live online coaching with AI specialists and produce custom video courses, so a program can be blended with in-person sessions or delivered fully online. We recommend in-person for practice-heavy courses and a blended format for organizations with many branches or shift work.
How is the cost of AI training set?
It depends on headcount, hours, format (talk, in-person practice, online, bootcamp), and whether we supply practice accounts. We do not quote from a fixed price list. After a free AI assessment and a consultation we propose a program and an estimate together.
Do you train civil servants and public institutions?
Yes. Our public-sector courses cover official document automation, civil complaint handling, AI literacy, and executive briefings, and they run within public-sector security and lab requirements. We also handle cohort-based operation and result reporting.
Do you teach Korea's AI Basic Act and AI ethics?
Yes. We run separate courses on the AI Basic Act and on AI ethics guidelines for management. They sit at the strategy and governance level, aimed at executives and planning teams setting the organization's rules for AI use.
Can we train on a single tool, such as Claude or Gemini?
Yes. ChatGPT, Claude, Gemini, Copilot, and Upstage each have a 16-hour intensive course. If your organization has already chosen a tool, take the intensive. If it has not, the role-based courses are the better fit.
Can we book only an instructor?
Yes. When an institution runs its own program and needs an AI or ChatGPT instructor, we place one. From 535 instructors in our network, we assign people who have passed all three checks — real implementation, institutional teaching, and one-to-one coaching — matched to the topic and audience.
Is the training open to small businesses?
Yes. There is an AI course for small and midsize companies and a one-day corporate class, so small teams can start without a large commitment. We also discuss AI talks for small-business associations.
How do you measure the results?
We assess AI use before the course, keep the work products participants build from their own tasks during it, and follow with a results report and adoption support. We judge by what actually moved into daily work, not by attendance or satisfaction scores alone.
Not the place you learn,
the place you work changes.