The story behind Wen

We didn't set out to build another AI SaaS product. We built Wen — a friend for the 3 million kids caught between two languages.

Here's the case for why that gap exists, why now is the moment to close it, and who's building it.

The problem

3 million Chinese-American kids
stuck between two languages

Their parents speak Chinese at home. Their schools teach in English. No existing app was built for this in-between world.

$80
Per hour for a Mandarin tutor
Most immigrant families can't afford private bilingual instruction. The gap widens every year.
2.3×
Higher literacy outcomes with bilingual tools
Research shows children who maintain heritage languages outperform peers in reading comprehension long-term.
68%
Of immigrant parents worried about language loss
In our interviews with 11 Bay Area families, every single parent cited this as a top concern before age 10.
Live product demo

Play it yourself. This is what a 6-year-old sees.

Every prompt is bilingual. Wen reacts in real time, just like on the homepage. Parents see detailed learning data after every session.

🧠
AI adaptive difficulty
Wrong twice on 山? The engine serves it again in 8 minutes, then adjusts the whole session trajectory.
🇨🇳 🇺🇸
Simultaneous bilingual exposure
Every question, every reward, every hint — delivered in both languages. Not sequential. Simultaneous.
📊
Parent insight dashboard
After each session: which characters need review, accuracy trends, and AI-generated homework suggestions for tonight.
learntowen.com/play
⭐ 0
Traction & market

Early signals. Massive market.

K–12 EdTech is a $295B global market. The bilingual heritage-learning niche is underserved and growing.

11+
Parent interviews completed. 9 of 11 said they'd pay $9.99/mo on first call.
3×
Game modes built and playable. Full bilingual content system with 10 base characters.
$2.8B
US bilingual children's education market. Growing 18% YoY driven by AI + immigrant demographics.
$9.99
Per month — below every comparable app. $120/yr vs $80/hr tutors. Parents understand this immediately.
Why now

Three forces converging

🤖
LLMs make real-time adaptation possible
Two years ago, per-child adaptive content required expensive human tutors. LLM APIs now make it possible at $0.001 per session. The margin structure just changed.
👨‍👩‍👧
Post-pandemic bilingual anxiety is peaking
Heritage language loss accelerated during COVID school closures. Immigrant parents are actively searching for solutions right now — CAC is at its lowest.
🏫
Schools are not solving this
Most US schools offer zero Chinese instruction. Chinese-American parents are buying this as a private service — the channel already exists, the product didn't.
Competitive position

No one else is doing all four

Product Chinese + English AI adaptive Parent dashboard Under $10/mo
Wen ✓ ✓ ✓ ✓
Duolingo Kids ✕ ✕ ✕ ✓
Lingokids ✕ ✕ ✕ ✕
Khan Academy Kids ✕ ✓ ✓ ✓
Studycat Chinese ✓ ✕ ✕ ✕
Founder

Built by someone who lived this

Founder-market fit isn't a talking point here — it's the origin story.

T
Tina
Founder & CEO, ATOP AI
Software engineer with full-stack + AI background. Chinese immigrant, Bay Area-based, bilingual in Mandarin and English. Built production systems at Sanhua (Flask, PostgreSQL, Docker, AWS, CI/CD). Spent the last decade building production software systems across the US and China — bringing a rare combination of deep technical skill and lived experience as a Chinese immigrant in America. ATOP AI is a way of giving the next generation a better start.
Flask + FastAPI PostgreSQL AWS LLM APIs Mandarin / English
V
Vignesh
Founder & CTO
Experienced in platform engineering, BD manager, and product management at various Fortune 500 companies, Vignesh leverages his diverse set of work experiences in his capacity as a founder. He has studied Mandarin as a non-native speaker, and is incredibly passionate about making language learning fun and engaging for all.
Product BD & Partnerships Platform Engineering Mandarin Learner