Vietnam · Conclusion 91 — English as a second language

An AI that teaches every Vietnamese child to read English.

Built at VinUniversity. Grounded in the Science of Reading. Designed to raise global citizens — responsibly, not just efficiently.

The national goal

Vietnam chose English. The destination is global citizens.

Conclusion 91 (2024) set the course: English a second language in every school — compulsory from Grade 1 by 2030, nationwide by 2035. The hardest, most decisive stage is the youngest: teaching children to read.

23M+
Students in Vietnam's school system
2030
English compulsory from Grade 1, nationwide
2035
English as a second language for all students
№ 91
Politburo Conclusion 91-KL/TW (Aug 2024)
The gap

The ambition is set. The teachers aren't there.

A nation can legislate English. It cannot hire fluent reading teachers fast enough to deliver it.

23,000,000
students who need to learn to read English
vs
30,000
teachers who specialise in English (of 1.05M total)

Vietnam needs 22,000 more English teachers by 2030, with ~120,000 teaching posts already unfilled. You can't hire your way to a nation that reads English — you have to scale the method.

The method

We don't guess at curriculum. We follow the science 42 US states made law.

The Science of Reading is the settled, evidence-based consensus on how children learn to read — five pillars confirmed across decades of cognitive science. Since 2013, 42 states and DC have written it into law. Our K–7 English line is built to that exact sequence.

i.

Phonics

Systematic decoding — the first online program built on UFLI (University of Florida) standards, 128 cumulative lessons.

Decode · Word recognition · K–3
ii.

Cedar Park

The bridge from decoding to fluent reading — vocabulary, accuracy, and automaticity.

Fluency · Vocabulary · transition
iii.

ASCEND

The Grade 4–7 English program — comprehension, background knowledge, and verbal reasoning.

Comprehend · Language comprehension · 4–7

Our product line is the science's progression — decode → fluency → comprehend. Vietnam adopts a legislated global standard, localised — not an experiment.

Responsible AI

AI that nurtures learning — not just optimises it.

The difference between AI that helps a child and AI that merely drills them isn't the model. It's the choices behind it.

"What separates beneficial AI from harmful AI is not the technology — it is human choices." — the responsible-AI philosophy Geoffrey Hinton calls for.

Our engine is grounded: it retrieves real curriculum before it answers, so it doesn't hallucinate. It reads a child's anxiety and attention, not just their score. And it measures real learning, not vibes.

Led by Simon Park — a pioneer of AI in education. We treat AI as a patient teacher, not a vending machine: deployed with care, transparency, and the child at the centre.

1
Grounded by RAG over a curriculum knowledge graph — no hallucinations, traceable answers
2
Error-driven — every mistake is classified and fixed, not papered over with more drilling
3
Psychology-aware — adapts difficulty, length, and tone to the child's cognitive state
4
Learning science — active recall, spacing, interleaving, deliberate practice, built in
5
Multi-stakeholder — students, teachers, and parents see the same honest picture
The combination

Vietnam's own AI. The world's proven English.

This isn't a foreign product sold into Vietnam. A Vietnamese university builds the engine; we supply the English it teaches. Neither half teaches a nation alone — together they do.

The engine

VinUniversity AI

The AI engine is built at VinUniversity — Vietnam's world-class research university (QS 5-star). Sovereign, homegrown deep tech.

×
The content

IPE English

Science-of-Reading curriculum, UFLI-based phonics, academy-audited question banks, and a proven per-language localisation playbook.

=
The result

A national English engine Vietnam owns

The scarce half is the content — anyone can attempt an AI tutor; few can supply English that actually teaches reading, localised for Vietnamese learners.

The journey

From a child's first English word to a global citizen.

One engine carries a learner the whole way — each stage powered by a product already running today.

AGE 6 · READS
Learns to read English
Systematic phonics on the Science of Reading. Reading Lab — Phonics · Cedar Park · ASCEND
AGE 12 · THINKS IN ENGLISH
Reads to learn — and builds proficiency
Comprehension, language, and exam readiness. EPT · 한 KOREAN-style localisation engine
AGE 16 · COMPETES GLOBALLY
Masters high-stakes assessments
Neuro-adaptive, learning-science SAT prep. AspireX — the AI engine, proven on the hardest test
BEYOND · BELONGS TO THE WORLD
A confident global citizen
Guided into universities and opportunities worldwide. IPE Big Picture Framework — consulting
Proof

Not deck-stage.

Real academies run on our software every day. Content is audited and already shipping internationally. The AI engine is operational.

3
Academies live on our operations platform
2,143
Audit-verified prep questions in our bank
128
UFLI-based phonics lessons, online-first
3
Languages live — EN · 한국어 · 中文

Academies in production

Ivy Academy, K Learning, and Reading Town Vancouver run daily on our platform — tri-lingual EN / 한국어 / 中文.

Content audited three ways

Every prep question triple-audited (Gemini Vision + Claude + human reviewer) against the College Board Bluebook style.

Localisation already underway

한 KOREAN ships per-language editions to institutes today — the same machinery localises English reading for Vietnamese learners.

Team

Operators who teach. Teachers who ship.

Classroom expertise, deep technical execution, and the discipline to run a real business while building software.

SP
Simon Park
AI & Learning-Science Lead · a pioneer of AI in education

Ph.D., P.Eng. · UBC PhD. 21 years at University of Calgary; VinUniversity; Direct-C. Leads the AI engine and the ethics-first approach to putting AI in the classroom.

PS
Paul Son
Director · IPE

Yonsei BA · UBC PhD. CEO of Readingtown Canada and Ivy Academy — the networks that make our products real businesses. Lead coach behind the EPT rubric.

DK
Daniel Kim
System Architect · Operations & Programs

Yonsei BBA · SFU MA Economics. Ex-PwC Seoul; built K Learning. Five published books. Owns the product — every workflow, every program, end to end.

Engineering & advisors

Engineering — Hanoi
Thuy Trang Nguyen (Dev · VNU Hanoi) · Duc Manh Doan (Dev · VNU Hanoi) · Brian Kim (PM, Calgary ↔ Hanoi)
MVP Team — Calgary
Andres Chacon (Dev · SAIT) · Kevin Hoang (Dev · UCalgary) · Olivia Yi (Intern · UWaterloo)
Editorial: Soonho Song, Ph.D. (Syracuse) — Editor-in-Chief, 한 KOREAN series.
Partnership

We're ready to pilot with Vietnam.

A public–private pilot that augments teachers rather than replacing them, measures real reading gains, and keeps data in-country. We've built BC-PIPA-compliant systems before — Vietnamese data residency is a known path, not a new one.

ModelPublic–private pilot · teacher-augmenting
PilotProvince cohort · measurable reading benchmarks
LocalisationVietnamese L1 edition · in-country data residency