Build an AI Startup in 10 Days

A 10-day practical program that uses AI to implement ideas into actual POC-ready services.

The ACE Startup program designs startup projects using AI software, We aim to provide product creation experience that extends from screen planning to prototyping, development, distribution, and demo.

View learning content View final output

Program Focus

AI Startup POC Product UX / Wireframe Prototype MVP Build Deployment Demo Pitch

Program Overview

This page is an HTML structure to introduce the AI-based start-up SW curriculum focusing on 10-day learning content.

10 Days
Intensive product development process
1 POC
Goals for actual, actionable results
AI + SW
Design and implementation of AI-utilized services
Team Build
Collaboration-based practical project implementation

10-Day Learning Contents

If you click on each card, you can check the detailed curriculum (session structure, learning goals, expected output, etc.) for each day for 10 days.

DAY 1
Day 1

Codex requirements specification & Figma screen design

Convergence practice to learn the basics of Figma and automate the requirements specification and reporting system using Codex Client

Starting from the original unstructured meeting minutes, we proceed with functional specification (FR), normal and exception flow definition, and Gmail-linked sending through Codex Client, and at the same time design Figma screen wireframes. This is day 1 learning content to understand the basics.

targetPlanner, developer beginner, prospective entrepreneur
What you needLaptop, Codex Client, Figma account
learning objectivesCodex AI planning automation practice and learning the basics of service screen design using Figma
2 hours Understanding Codex AI planning guide and Figma basic interface
3 hours Functional Requirements (FR) Table and Exception Flow Design Practice
1 hour Gmail plugin actual sending and sharing of results
Requirements Codex AI Gmail Plugin Figma Grid Wireframe
expected output requirements-specification.md, final shared email in sent mailbox, Figma screen basic draft
DAY 2
Day 2

Sequence Diagram & Data Flow Design

Data flow design practice connecting screen design to system operation flow

User behavior, data flow, and system interaction are based on Figma screen design. This is the second day of learning content that visualizes with a sequence diagram and even designs exception situations.

key topicsSequence Diagram, Data Flow, System Interaction, Exception Handling
learning objectivesUnderstand and design interactions between user-client-server-DB in chronological order
Practice directionExtend the system perspective by converting the screen structure into function flow and data flow
2 hours Understand sequence diagram and data flow concepts
3 hours Core Features Scenario Writing and Diagram Practice
1 hour Share results and review exception handling
Sequence Diagram Data Flow System Interaction Exception Handling Hands-on
expected output Core functional scenarios, data mapping table, sequence diagram, exception response matrix
DAY 3
Day 3

Data Normalization & ERD Design with Codex AI

Practice converting screens and user flows into data structures and ERDs and using Codex AI to refine the design

Codex Client Workspace is linked to the process of extracting necessary data based on screen design and user flow and defining ERD through normalization. Inconsistency in data understanding between planning and development Resolution (Task 01), construction of a project-wide terminology standard dictionary (Task 02), analysis of API/UI chain impact when modifying schema (Task 03), and logging of design decision history (Mission 01), Perform concurrency/integrity error exception case discovery (Mission 02) at a practical level.

key topicsData Extraction, Entity Definition, Relationship Design, Normalization, ERD, Codex Workspace
learning objectivesDesign a data structure that fits the screen and flow, and use Codex AI to perform normalization and precise ERD verification.
Practice directionScreen → User flow → Data → ERD connection and Codex-based design automatic auditing
2 hours Understanding ERD design and aligning data structure design (Task 01)
3 hours Terminology standardization (Task 02) and schema change impact scope analysis (Task 03)
1 hour Design decision logging (Mission 01) and exception scenario discovery (Mission 02)
ERD Normalization Codex AI Glossary Impact Analysis Decision Log Edge Case
expected output ERD diagram, normalization analysis document (erd-schema.md), standard terminology dictionary (project-glossary.md), change impact report, design decision history (DDL), exception scenario analysis
DAY 4
Day 4

API Design & Swagger Specification with Codex AI

[Codex AI linkage] Convert screens and data structures into actual developable API specifications and use Codex AI to Practice auditing interface contracts and consistency

In the process of defining actual service functions in API form based on screen design, user flow, and data structure and documenting them in Swagger/OpenAPI format. Codex Client WorkspaceConnect. Automatic creation of AI-based API contract (Task 01), detection of inconsistency between planning specification and actual API structure (Task 02), API specification Analysis of UI/DB chain ripple effects upon change (Task 03), autonomous audit of RESTful API standard specifications (Mission 01), and boundary value edge case test scenario design (Mission 02) Perform at a practical level.

key topicsAPI Design, Request/Response, Swagger, OpenAPI, Codex Workspace, Consistency Audit, Impact Analysis
learning objectivesDefine API specifications based on screens and data structures and link Codex AI to ensure consistency and precision of REST specifications Conduct Auditing
Practice directionScreen ➔ Flow ➔ Data ➔ API ➔ Swagger Structure Design Connection and Codex Client Autonomous audit of base design defects
2 hours Understanding API concepts and automatically generating Codex-based API contracts (Task 01)
3 hours Consistency check between planning and API (Task 02) and auditing the scope of impact of API changes (Task 03)
1 hour Autonomous review of RESTful specifications (Mission 01) and creation of edge case test scenarios (Mission 02)
Codex AI API Design Swagger OpenAPI Consistency Audit Impact Analysis API Review Test Scenario
expected output API contract (api-contract.md), API consistency report (api-consistency-report.md), change impact analysis (api-impact-analysis.md), REST Specification review report (api-review-report.md), edge test scenario (api-test-scenario.md)
DAY 5
Day 5

UI/UX Design & Interaction Design with Codex AI

[Codex AI linkage] Auditing user friendliness and interface usability with Codex AI based on wireframes UI/UX review practice to ensure quality

Based on the constructed screen plan and user movement, we detect objective cognitive friction points and supplement UI consistency. Codex Client WorkspaceConnect. Usability review to analyze whether users can easily understand (Task 01), audit of UI specifications and terminology consistency between multiple screens (Task 02), Optimize journey flow by shortening depth (Task 03), collect trouble logs by applying persona for the elderly underprivileged (Mission 01), and evaluate comprehensive service quality based on multidimensional ratings (Mission 01) 02).

key topicsUI/UX Review, UI Consistency, User Journey Map, AI Persona, Usability Score, Codex Workspace
learning objectivesObjectively analyze the usability of the planning screen and utilize Codex AI to provide intuitive UI/UX and shortened flow path Derivation of optimization plan
Practice directionWireframe ➔ UX Audit ➔ UI Consistency ➔ Journey Optimization Structural design and Codex-based usability quality assurance
2 hours User perspective screen review (Task 01)
3 hours Review consistency between screens (Task 02) and improve user movement (Task 03)
1 hour Persona usability review (Mission 01) and comprehensive service quality evaluation (Mission 02)
Codex AI UI/UX Review UX Flow Friction Audit UI Consistency User Journey AI Persona Usability Score
expected output UX review report (ux-review-report.md), UI consistency report (ui-consistency-report.md), movement improvement report (user-journey-report.md), first User review (first-user-review.md), comprehensive service quality evaluation (usability-evaluation.md)
DAY 6
Day 6

Front-End Development & API Integration with Codex AI

[Codex AI linkage] Development practice that implements Figma screens into actual operating service screens, debugs with Codex AI, and connects API data.

Implementing Figma UI as an actual service screen, binding server data to the screen through fetch/AJAX-based API integration, and handling loading, errors, and empty states throughout the entire development process. Codex Client WorkspaceConnect. Build a workable screen draft based on the screen specification (Task 01), read and defend code patches for console errors such as runtime TypeError (Task 02), refactor to remove duplicates and organize them into readable variable names (Task 03), build a report to troubleshoot CORS policy and JSON parsing errors (Mission 01), and review and encapsulate code quality such as blocking global contamination (Mission 02). Perform.

key topicsPrototype Generation, Stack Trace Debugging, Defensive Coding, Code Refactoring, Error Troubleshooting, Code Quality review, Codex Workspace
learning objectivesCollaborate with Codex AI pair programmers to quickly build screen prototypes from blank screens, resolve console errors, organize highly readable source code, and understand API integration.
Practice directionFigma UI ➔ Build screen drafts ➔ Error debugging ➔ Code refactoring ➔ Exception response and encapsulation quality assurance
2 hours Quickly create a screen draft (Task 01)
3 hours Error Resolving Assistant (Task 02) and Improving Code (Task 03)
1 hour Fix errors with AI (Mission 01) and get AI code reviews (Mission 02)
Codex AI Prototype AI Debug TypeError Guard Refactoring Troubleshooting Code Review IIFE Encapsulation
expected output Responsive dashboard (index.html, layout.css), debugging script (day6_debug_solution.js), refactoring source code (day6_clean_code.js), debugging result report (day6_error_report.md), final audit result report (day6_code_review.md)
DAY 7
Day 7

Development: Completing Features with Codex AI

[Codex AI linkage] Function completion practice that connects screens and functions, checks user registration and reservation scenarios, and verifies and blocks abnormal input exceptions.

Beyond the implementation of the Figma screen skeleton, the development process involves tying event mapping and data storage to the static UI so that actual actions are performed, and ensuring operational stability for chained scenarios. Codex Client WorkspaceConnect. Assemble screen markup event connection and local storage loading (Task 01), preservation guard to maintain filter settings when moving to multiple screens (Task 02), prevention of duplicate registration and verification of past date blocking (Task 03), quality rating for missing loading feedback (Mission 01), and E2E user simulation driving verification report (Mission 02).

key topicsFeature Integration, JavaScript Event, LocalStorage, User Flow review, SessionStorage, Debounce Lock, Logic Validation, Codex Workspace
learning objectivesCollaborate with Codex AI scenario coaches to assemble highly complete service functions that map events to static markup, ensure data preservation, and completely control abnormal input errors.
Practice directionScreen Skeleton ➔ Function Mapping ➔ Journey Preservation ➔ Multi-hit Blocking Exception Guard ➔ E2E Completeness Auditing
2 hours Connecting functions (Task 01)
3 hours Verifying user scenarios (Task 02) and finding functional errors (Task 03)
1 hour Service completeness inspection (Mission 01) and E2E simulation driving verification (Mission 02)
Codex AI Integration Event Map LocalStorage State Backups Validation Submit Lock E2E Verification
expected output Feature integration page (day7_connected_feature.html), session restoration script (day7_scenario_solution.js), security guard completion (day7_robust_validation.js), completion report (day7_service_audit.md), scenario validation (day7_e2e_validation.md)
DAY 8
Day 8

User Management: Membership & Onboarding UX

[Codex AI linkage] From member service design to onboarding, review of friction in the user journey and practice of implementing redundant sign-ups and security-vulnerable exception guards

We are not limited to implementing the login technology module, but we also aim to optimize the sign-up process from a user convenience perspective and solidify account consistency. Codex Client Workspaceclosely coordinated. We build a priority plan for acquiring essential member functions (Task 01), a flow design to remove distribution and exit barriers for each step of the input form (Task 02), a duplicate signup blocking consistency verification and email format Regex guard coding (Task 03), a mock signup emotional feedback report for silver customers (Mission 01), and a quality evaluation report containing a hacking brute force prevention information hiding patch (Mission 02).

key topicsUser Feature Planning, Onboarding UX, Form Friction Audit, Progressive Profiling, Database Integrity, Duplicate Account Guard, Info Hide, Usability Evaluation
learning objectivesWork with Codex AI user experience coaches and audit partners to assemble a user-friendly member experience that minimizes sign-up friction, ensures account integrity, and blocks security exceptions.
Practice directionFeature derivation ➔ Barrier review ➔ Storage integrity guarding ➔ Beginner virtual experience ➔ Rating and security assessment
2 hours Member function design (Task 01)
3 hours User form review (Task 02) and data consistency guard coding (Task 03)
1 hour Silver Persona Onboarding Auditing (Mission 01) and Error Message Concealment Security Patch (Mission 02)
Codex AI Membership UX Audit Integrity Email Regex Duplicate Guard Usability Evaluation
expected output Feature Priority List (day8_feature_checklist.md), Onboarding Improvement Report (day8_ux_improvement_report.md), Anti-Duplicate Guard Code (day8_robust_user_service.js), First User Review (day8_first_user_review.md), Quality Evaluation (day8_usability_evaluation.md)
DAY 9
Day 9

Payment Service: Experience & Trust Audit

[Codex AI linkage] User-centered payment journey simplification design and financial exception scenarios Safe processing and reliable UI auditing in English

Beyond the success of technology-centered PG API calls, we aim to enable users to make payments with peace of mind without any anxiety. Codex Client WorkspaceUse closely. Planning a 3-step or lower payment flow that merges multi-step exit flows (Task 01), converting error codes such as card limit exceeding/delay into Korean and designing to block double billing (Task 02), importing receipt price segmentation discounts and security communication badges (Task 03) and inspection time waiting messages (Mission 01), and building a journey diagnosis report (Mission 02) to correct coupon search form loss and peak friction.

key topicsPayment Flow Planning, Trust Auditing, Exception Playbook, Friendly Error Mapping, Receipt Breakdowns, Idempotency Guard, Friction Peak Auditing
learning objectivesCollaborating with Codex AI payment flow consultants and quality auditors to plan a safe payment structure that dramatically reduces the user payment abandonment rate and establishes feedback to resolve financial failures.
Practice directionMerging of routes ➔ Error replacement ➔ Safety badge injection ➔ Failure resolution response ➔ Journey peak correction
2 hours Payment flow design (Task 01)
3 hours Review exception scenarios (Task 02) and improve secure payment UI (Task 03)
1 hour Writing a Disability Reassurance Statement (Mission 01) and E2E Journey Peak Audit (Mission 02)
Codex AI Payment UX Trust Audit Friendly Error Idempotency Friction Peak Mitigation Playbook
expected output Simplified order payment flow (day9_optimized_flow.md), safe error phrase book (day9_error_mitigation_guide.md), revised safe payment UI (day9_robust_payment_ui.html), failure response statement (day9_failure_scenarios.md), customer journey diagnosis (day9_checkout_journey.md)
DAY 10
Day 10

Service Launch: Launching a service

Completion of the final task of the actual service planner, from creating a service introduction to preparing user support, organizing project results, and establishing a Version 2.0 roadmap.

It is the starting point of service, not the end of development. With Codex AI, you will experience the final work flow of an actual service planner: writing a launch introduction and opening notice, predicting difficulties that users will experience, preparing FAQs and customer service guides, and establishing a project results report and Version 2.0 roadmap.

key topicsLaunch Prep, FAQ, Customer Support, Retrospective, MoSCoW, Roadmap, Presentation, Sprint Plan
learning objectivesEstablishment of user launch package, performance report, and Version 2.0 planning roadmap
Practice directionLaunch prep → Customer Support → Performance Reporting → Version 2.0 Roadmap
2 hours Service launch preparation and user support design (Task 01 & Task 02)
2 hours Summary of project performance and Version 2.0 roadmap (Task 03)
2 hours Launch package completion and final announcement (Mission 01 & Mission 02)
Launch Prep FAQ 20 KPT Retrospective MoSCoW Roadmap Final Presentation
expected output Launching materials (day10_launch_materials.md), FAQ & response guide (day10_faq_guide.md), result report & roadmap (day10_retrospective.md), integrated launching package (day10_launch_package.md), final presentation material (day10_presentation.md)
PROJECT DAY
Project Day

Markdown Spec-Based Website Building (Project Day)

A 6-phase hands-on workshop to structure ideas into Markdown spec files via Codex conversation and build a fully functional website

Experience Idea Definition (Phase 1), 5 Essential Markdown Spec Writing (Phase 2), Codex Website Generation (Phase 3), Student & Teacher Scenario Testing (Phase 4), Spec-First Refinement (Phase 5), and Conversation History Logging & 5-min Demo (Phase 6).

Core TopicsIdea Definition, Markdown Specifications, Codex Implementation, Scenario Testing, Spec-First Refinement, History Logging
Learning GoalsMastering the process of collaborating with Codex AI based on standard specifications to launch working web applications
Practice FlowIdea ➔ Specs ➔ Codex Build ➔ Scenario Test ➔ Spec Refinement ➔ History & Demo
30 min Idea Definition & 5 Markdown Specs (Phase 1 & 2)
45 min Codex Code Generation & Role Testing (Phase 3 & 4)
45 min Spec & Code Refinement, History Logging & Demo (Phase 5 & 6)
Project Day Codex Website Markdown Specs Scenario Test Spec-First CONVERSATION.md
Expected Deliverables Site Specification (SITE_SPEC.md), User Flows (USER_FLOWS.md), Data Model (DATA_MODEL.md), Implementation Plan (IMPLEMENTATION_PLAN.md), Conversation Log (CONVERSATION.md), Complete Website (website/)
ADD LECTURE
Add LECTURE · Thu/Fri Sprint

AI Startup Sprint × GStack (Thu/Fri Two-Day Sprint Deck)

Codex app + GStack — 2-day sprint deck covering problem interrogation, scope cutting, 1-screen demo, and talk design (95 slides)

Explore 95 slides covering hypothesis formation, evidence gathering, GStack workflow skills ($office-hours, $plan-ceo-review, $design-html, $review, $qa), and Friday demo day presentation structure.

Core TopicsHypothesis Interrogation, Scope Reduction, Evidence Collection, Demo Path, Talk Design, GStack Skills
Learning GoalsMastering two-day AI startup sprint workflow and evidence-first product judgment
Add LECTURE GStack Codex App Thu/Fri Sprint 95 Slides
ADD LECTURE
Add LECTURE · Student Workbook

Thu–Fri Student Workbook — Build It in One Day, Present It the Next

Comprehensive Thu/Fri student workbook with worksheets, evidence logs, 5-slot script templates, and prompt pack P1–P8

Includes 4 core deliverables (hypothesis, 2 evidence records, 1 demo path, 5-slot talk script), 6 GStack skill catalog, installation & zero-install prompt pack guide.

Core TopicsWorksheets A1-D1, Prompt Pack P1-P8, Evidence Ladder, Demo Day Rubric, Troubleshooting
Learning GoalsHands-on workbook for executing Thursday sprint and preparing Friday 5-minute presentation
Add LECTURE Workbook Prompt Pack P1-P8 Rubric Thu/Fri Sprint
ADD LECTURE
Add LECTURE · Student Supplement

Superpowers — A Beginner's Guide (Supplementary)

Free 14-skill open-source pack teaching AI disciplined software engineering: design first, plan, build, systematic debug & verify

Optional supplement for student personal projects. Includes 1-click Codex app install, 6 core skills ($brainstorming, $writing-plans, $executing-plans, $systematic-debugging, TDD, $verification), workflow loops, and Superpowers vs GStack guide.

Core TopicsBrainstorming, Writing/Executing Plans, Systematic Debugging, TDD, Verification, 1-Click Plugin Install
Learning GoalsMastering 14-skill AI software methodology for building disciplined post-sprint personal projects
Add LECTURE Superpowers Codex Plugin 14 Skills Supplement
ADD LECTURE
Add LECTURE · Resource Drive

LECTURE Drive — TOT AI Shared Educational Resources

Sharing educational materials and resources used in TOT AI.

Access and download full presentation slides, practice workbooks, prompt guides, and supplementary reference materials used in the TOT AI course directly via the Google Drive shared folder.

Core TopicsTOT AI Courseware, Practice Materials, Shared Google Drive
Learning GoalsConvenient access to all teaching and learning resources for review
Add LECTURE LECTURE Drive TOT AI Google Drive Resources
SURVEY & GRADUATION
Final Survey · Completion Notice

All classes in the AI TOT Track have ended. Thank you.

All classes in the AI TOT Track have ended. Thank you. Please complete the final survey.

Congratulations on completing the course! Wishing you endless success and happiness in your future endeavors. Click below to view the graduation message and QR codes for the 2 final surveys.

Key NoticePost-Course Survey & AI Track Pre-Learning Diagnosis QR Codes
PurposeCourse evaluation feedback & learning competency assessment
Final Survey Graduation QR Codes Feedback Assessment

Final Outcome

The goal of this course is not simply to learn, but to complete a POC product and demo flow that can be demonstrated in practice.

Wireframe
Core screen structure design
Prototype
Clickable user flow
POC Build
Demonstration of a product that can actually operate
Demo Pitch
Final announcement and expansion possibility review