APLANNER - AI-POWERED PRODUCTION PLANNING · MVP 2026 · FOUNDING AMOZA KOREA 2027

AI for Real Life.
Real Impact.

Amoza builds AI for the value chain — from shop floor to customer door. Today we build step one: aPlanner, an AI-powered production-planning engine for semiconductor and high-tech manufacturing, created by the founder in Korea.

MAKE > MOVE > SELL > DELIGHT — a direct-to-customer business model where production, delivery, selling and customer care run on one AI platform. Plan Smarter. Produce Faster.

1
Product in MVP (aPlanner)
36
Product roadmap
7
Pillars, one platform
2027
Amoza Korea Co., Ltd. incorporation (TBC)
Built for:
Semiconductor
Electronics
Automotive
Heavy industry
WHERE WE ARE TODAY — OCTOBER 2026

An honest picture of the stage we are at

Pre-commercial MVP

aPlanner prototypes run on synthetic fab data. Phase 3 "Validate" (Oct–Dec 2026) is in progress; the first Korean fab pilot is planned for 2027. No customers, no revenue yet.

Founder-led, no legal entity yet

Personal research by the founder, Nguyen Van Duong (Danniel Ng), self-funded and supported by two Siemens Industry Software advisors and five engineering partners.

Founding in Korea

Amoza Korea Co., Ltd. (주식회사 아모자코리아) — incorporation Mar–May 2027, Incheon Songdo or Suwon (TBC), under the Technology Startup Visa (D-8-4) path with OASIS.

STEP ONE — IN MVP

aPlanner - AI-Powered Production Planning

Plan Smarter. Produce Faster. Four AI modules, one planning engine — built for semiconductor complexity: yield-aware, constraint-based, scenario-driven.

MAKE
MVP

aPlanner

AI-Powered Production Planning

Korean fabs make the world's memory, but they still plan it in Excel: 3–5 days per planning cycle, 55–65 % forecast accuracy, ~70 % of planner time spent gathering data. aPlanner is designed to cut planning cycles from days to hours — on-premise or in a private cloud, so production data never leaves the fab.

AI Demand Forecasting
Target 85–90 % accuracy (+20–30 pts vs. spreadsheets)
Intelligent Scheduling
Target: full-fab schedule in < 30 min, 99 %+ feasible plans
Scenario Simulation
Target: minutes per what-if scenario, side-by-side KPIs
Real-Time Analytics
One live source of truth; every recommendation explainable
Target planning cycles from days to hours — MVP under validation on synthetic fab data (2026)
TARGET OUTCOMES — MODELLED ON SYNTHETIC FAB DATA

What aPlanner is designed to deliver

Every figure below is a design target from simulation on synthetic fab data. There are no customer deployments yet; pilot results are expected in 2027.

85–90 %
Forecast accuracy
Target vs. 55–65 % with Excel-based planning
< 30 min
Full-fab schedule
Target — 80 % faster than 3–5-day manual cycles
99 %+
Feasible plans
Target — constraint- and yield-aware scheduling
Minutes
Per scenario
Target — 95 % faster than days in spreadsheets
20 → 60 %
Planner time on strategy
Target — less data gathering, more decisions
8–12 wks
To a running pilot
Target vs. 12–18 months for legacy APS
THE ROADMAP

1 product in MVP · 36-product roadmap

A six-step staircase: aPlanner → aMonitor → Amoza AI Platform → Smart Factory (MAKE) → Smart CX (DELIGHT) → the complete chain with MOVE and SELL. Every step reuses the data and AI of the step before.

MAKE
Smart Factory
aPlanner MVP · suite 2029–31
aPlanner · MVP
aMonitor · Next
aOps
aQuality
aYield
aMake
Explore MAKE
AI FOUNDATION
Amoza AI Platform
In development · 2028–29
aCore
aFlow
aConnect
aExecutor
aData
Explore AI FOUNDATION
DELIGHT
Smart Customer Experience
Roadmap · 2031–33
aCare
aLoyalty
aInsight
aVoice
aCommunity
aPersona
Explore DELIGHT
MOVE
Smart Supply Chain
Roadmap · 2033+
aFulfill
aShip
aRoute
aStock
aReturn
aTrack
Explore MOVE
SELL
Smart Commerce
Roadmap · 2033+
aShop
aPrice
aSearch
aMarket
aConvert
aSocial
Explore SELL
DEVELOPER
Developer & Platform
Roadmap
MendixOne
aForge
aHub
Explore DEVELOPER

Everything starts with planning.

Start with aPlanner
AMOZA AI PLATFORM — IN DEVELOPMENT

One platform serving every pillar

Five components that aPlanner uses first and every later product reuses — one data model from production to inventory, orders and customer signals.

In development

aCore

ML models, training pipelines and inference — forecasting for the fab today; inventory, pricing and churn tomorrow.

In development

aFlow

Data pipelines and data quality — one ingestion layer for MES, ERP, WMS, commerce and CX data.

In development

aConnect

ERP · MES · commerce · CRM integration hub. Pre-built connectors: SAP, Oracle, MES — in development; read-only first.

Roadmap

aExecutor

Autonomous AI agents that run multi-step workflows across pillars — re-plan, re-allocate, notify the customer.

Roadmap

aData

Governance, lineage and PII detection — required for Korean fabs and for customer data alike.

Planned

Security roadmap

On-prem / private cloud first, RBAC, SSO, audit trail. Roadmap: SOC 2 Type II (2027–28), ISO 27001 (2028), Korean ISMS (planned).

VALIDATION

Validated on synthetic fab data — pilot results coming 2027

No customer quotes, no case studies — not yet. Here is where development stands and who is behind it.

Development stage

Phase 1 — Foundation
Apr–Jun 2026 · Done
Phase 2 — Concept
Jul–Sep 2026 · Done — prototypes on synthetic fab data
Phase 3 — Validate
Oct–Dec 2026 · In progress
Phase 4 — Launch prep
Jan–Mar 2027 · Planned
First Korean fab pilot
May–Sep 2027 · Planned — 8–12 weeks, one line

Who is behind it

Founder & Representative
Nguyen Van Duong (Danniel Ng) — Senior Software & AI Engineer; Industry Software and AI projects for Siemens Industry Software and Samsung Electronics in Korea. Originator of aPlanner.
Advisors — Siemens Industry Software
Two senior architects advising on Industry Software architecture, MES / PLM integration and Korean customers.
Partners
Five engineering partners (AI/ML, full-stack, applied AI, product/UX, finance) working under the founder’s direction.

No legal entity yet — personal research by the founder. Amoza Korea Co., Ltd. incorporates in 2027.

AMOZA KOREA

Founded in Korea, for Korea's fabs

Korea is the center of gravity of memory manufacturing. Amoza Korea Co., Ltd. will hold the IP, do the core R&D and sign every contract — with the founder in Suwon, Korean hires in 2027, Korean UI and KRW pricing. Data stays in the fab.

Incorporation 2027
Songdo or Suwon (TBC)
OASIS · D-8-4
FOR DEVELOPERS — ROADMAP

Build, extend, integrate — later

The developer platform — low-code with MendixOne, no-code with aForge, APIs via aHub — follows the Amoza AI Platform (2028–29). Today the integration work is read-only connectors for MES, SAP, yield databases and Excel.

MendixOne
aForge
aHub
See the developer roadmap

Pilot program 2027 — Korean fabs & suppliers

Become a design partner and put AI planning into your fab — from shop floor to customer door, starting with the plan.

Paid Planning Readiness Assessment first
8–12-week pilot on one production line
On-prem / private-cloud deployment — data stays in the fab
Korean UI, Korean support, KRW pricing
aPlanner pricing (KRW / USD)

Pre-commercial MVP · no customers yet · founder-led · Amoza Korea Co., Ltd. incorporating 2027.