mew.monster

Software & AI education company

MEW Monster뮤몬스터

We build software that makes everyday work easier.

We are MEW Monster. We build tools for music academies and product teams, and create AI learning experiences. From daily operations to a first AI project, our products help people put technology to work.

Meet MEW Monster Explore our products
DevOps · Schedule
Modu AI Lounge
Monster SaaS · Revenue

Who we are · MEW Monster

One team. Connected possibilities.

MEW Monster is a software company connecting operations, product development and AI education. Our five products share one purpose: helping people use technology in their everyday work and learning.

01 · Operations

Closer to the way people work

With Monster SaaS, we bring lessons, attendance, practice rooms and settlement into one place for music academies.

Discover Monster SaaS → 02 · Product development

From an idea to a working product

With DevOps, we connect specifications, QA and project operations, and use Claude to help organize product knowledge.

Discover DevOps → 03 · AI education

Technology people can learn and use

Through Modu AI Lounge, AI Lab Lectures and API Checker, we connect AI learning with hands-on practice and public data.

Explore AI education →
5products
10,086public APIs cataloged
5partner media centers
1,900+automated tests in DevOps
9academy workflows in Monster SaaS

Made by MEW Monster

Our work, in five products.

Academy SaaSMonster SaaS

Lessons, rooms, revenue and payroll for multi-branch music academies.

 Team platformDevOps

Specs, QA, issues, schedules and API health in one workspace.

 Data toolAPI Checker

Finds which of 10,086 public APIs really return data.

 Public siteModu AI Lounge

AI classes at five media centers, in one place.

 EducationAI Lab Lectures

Beginner courses from first prompt to automation.
Flagship · Academy management

Monster SaaS

Music academy operations

One screen for the front desk: today's lessons, attendance, payments and alerts. Behind it, practice-room booking, revenue and teacher payroll for every branch.

  • Academy directors
  • Branch managers
  • Front-desk staff
  • Teachers
Branch switcher

Every page follows the selected branch, or all branches at once.

One-tap attendance

Scheduled → present → late → absent, saved instantly.

Make-up lessons

Absences without a make-up booking show up as to-dos.

Practice rooms

Hourly occupancy grid, overlap checks, booking rules per room.

Revenue

12-month trend per branch, manual cash entries, CSV export.

Teacher settlement

Calculate, adjust with reasons, approve and lock, with a full audit log.

Monster SaaS · Teacher settlement

Demo data. Names, contacts and amounts are blurred.

Practice roomsRooms, hours, fees and booking rules
ConsultationsUnified inbox across channels
Students & guardiansSearch, filter, add, link to classes
Lesson calendarLessons, practice and make-up bookings
Team platform · Specs to release

DevOps

Feature specification & project operations

Planners, developers, designers and QA work from one source of truth. Every project gets its screens, flows, events, data model, test cases, issues, documents, schedule and API health in one workspace.

  • Product planners
  • Developers
  • Designers
  • QA
  • Project leads
Living specs

Screens, user flows, event specs and a data model that stay linked.

Flow map

Auto-laid-out map of screens and features, exportable to Markdown.

QA design

Test cases with draft → approve → retire and change tracking.

Schedules

Monthly calendar, task sheet and WBS import across projects.

API health

Uptime bars, incidents and reports for each service.

History

Revisions, baselines and side-by-side compare.

DevOps · Data model

Project content is blurred. Menus and layout are shown as-is.

User flowsFeatures with screens, actors and routes
Flow mapScreen and feature lenses
Test design1,316 cases from 149 scenarios
Global dashboardPriority tasks across projects
Data tool · Public open APIs

API Checker

API 확인소 · Digital Convergence Platform

Korea's Digital Convergence Platform lists over ten thousand open APIs. API Checker shows which ones actually return data, and which ones work inside the Modu AI Lab vibe-coding studio, with the reason for each result.

  • Educators
  • Program teams
  • Citizen developers
10,086 APIs, 18 fields

Full catalog with completeness checks on every refresh.

Real data, verified

A 200 response is not enough; empty and error payloads are told apart.

Studio tags

Restaurants, parking, tourism, weather and bus data ready for class.

Safe by design

Runs locally; keys stay in memory and never appear in URLs.

API Checker · Catalog

Verification details are blurred.

Public website · AI education

Modu AI Lounge

모두의 AI 라운지

Helps citizens, families, students and teachers find hands-on AI classes at five Gyeonggi-province media centers: Gimpo, Goyang, Seongnam, Hwaseong and Anseong.

  • Citizens
  • Families
  • Students
  • Educators
Interactive home

Learn, experiment, together: three ways in, with motion you can pause.

Course finder

50 courses filtered by center, shareable by link.

Center pages

Verified addresses, photos and each center's classes.

News

42 notices from all centers with search and filters.

Modu AI Lounge

Course and notice details are blurred on inner pages.

CoursesFilter by media center
Media centersFive partner centers
NewsNotices from every center
Education service · Modu AI Lab

AI Lab Lectures

모두의 AI 실험실 강의

Beginner lectures for the national Modu AI Lab platform. Every course follows one example: build a to-do app, connect a weather API, then send a morning reminder automatically.

  • Non-developers
  • Teachers
  • Media center learners
Vibe coding 1–3

Sign-in, first app from a prompt, real public data.

n8n automation

Schedule trigger, HTTP request, email at 8 AM.

Cloud & dev tools

Cloud VM and managed open-source tools.

Privacy mask

Our Chrome extension hides personal data while recording.

Title slides from the lecture decks.

Monster AI · powered by Claude

Claude connects Monster SaaS and DevOps.

Monster SaaS knows how academies run. DevOps knows how the product is specified and tested. With the Claude API in between, product knowledge stays organized automatically, and academy staff get plain-language help with their daily work.

People approve. Everything Claude writes lands in DevOps as a draft. A planner or QA lead approves it, and revision checks stop a draft from overwriting newer edits.

Organize

Specs from feature docs

Turn each Monster SaaS feature document into DevOps features, events and exception cases.

feature doc → spec draft
Organize

Drift check

Compare the DevOps spec with real Monster SaaS routes and tables, and list what is missing on either side.

spec export + routes → report
Quality

Test case drafts

Draft QA cases for booking conflicts, settlement locks and attendance changes.

spec → QA drafts
Operate

Incident summaries

Explain what failed, when, and the likely cause from DevOps health checks.

health report → note
Operate

Release notes

Turn a history compare into notes for academy staff and a changelog for developers.

history compare → notes
Optimize

Academy insights

Flag unusual payroll adjustments, a growing make-up backlog or rooms that are always full.

totals → weekly summary
Assist

Reply drafts

Draft answers to parent inquiries from schedules and room policies; staff edit and send.

inquiry → reply draft
Organize

Data summaries

Group failed API checks by cause and explain them for the education team.

API Checker → report
spec-assistant.ts · Anthropic TypeScript SDK
import Anthropic from "@anthropic-ai/sdk";
import { z } from "zod";
import { zodOutputFormat } from "@anthropic-ai/sdk/helpers/zod";

const client = new Anthropic();

// Matches DevOps features, events and exceptions
const SpecDraft = z.object({
  features: z.array(z.object({
    name: z.string(),
    screen: z.string(),
    description: z.string(),
    exceptions: z.array(z.string()),
    events: z.array(z.object({
      code: z.string(),
      trigger: z.string(),
      output: z.string(),
    })),
  })),
  gaps: z.array(z.string()),
});

export async function draftSpec(featureDoc: string, specExport: string) {
  const res = await client.messages.parse({
    model: "claude-opus-5-5",
    max_tokens: 16000,
    output_config: { effort: "high", format: zodOutputFormat(SpecDraft) },
    system: [
      { type: "text", text: AUTHORING_GUIDE },
      // Large, stable spec export: cached across calls
      { type: "text", text: specExport, cache_control: { type: "ephemeral" } },
    ],
    messages: [{
      role: "user",
      content: `Draft DevOps features, events and exceptions for this
Monster SaaS feature doc. List gaps against the current spec.

${featureDoc}`,
    }],
  });

  if (res.stop_reason === "refusal") throw new Error("Claude declined");
  return res.parsed_output; // saved in DevOps as a draft
}

Safe and affordable by design

  • Drafts only. Claude never approves its own output.
  • Typed output. Every draft is checked against the DevOps schema before saving.
  • Through the API. Claude uses DevOps endpoints, so roles and revision checks still apply.
  • No personal data. Student and guardian details stay in Monster SaaS; prompts get counts and totals.
  • Low cost. The spec export is cached, nightly checks run as batches at half price, and effort is set per task.
  • Traceable. Each call is logged with its token usage next to the change it produced.
How we build

Built to run every day

Modern stack

React 19 and TypeScript front ends, NestJS and Next.js APIs, MySQL 8.4, Docker-based delivery.

Tested

DevOps alone runs about 1,665 unit and integration tests and 287 browser tests on every change.

Secure sign-in

Company SSO with OIDC and PKCE, server-side sessions, and role-based access per project.

Privacy first

Personal data stays on the server; dashboards get totals. Keys never go into URLs.

Accessible

Keyboard support, reduced-motion modes and automated accessibility checks on public sites.

AI-native team

We build with Claude Code under written rules and design Claude features with human approval.

For AI readers

Machine-readable company profile

AI agents can read this page directly. The same facts are published as structured data next to it.

portfolio.json: company, products, features, users and the Claude plan.

llms.txt: a short plain-text summary.

The page head also carries schema.org Organization JSON-LD.

COMPANY   뮤몬스터 (MEW Monster) · mew.monster · Korea
FOCUS     music academies · product teams · AI education
PRODUCTS
  Monster SaaS     academy operations SaaS (flagship)
  DevOps           specs, QA, schedules, API health
  API Checker      10,086 public APIs verified
  Modu AI Lounge   AI classes at 5 media centers
  AI Lab Lectures  vibe coding, APIs, n8n, cloud
AI        Claude API links Monster SaaS → DevOps
          drafts only, people approve