PROJECT FOOTAGE / HARRY
FEATURED / ROBOTICS & PLAY
DriverLabby Harry
A robot-driving game.
A whole arena of decisions.
3DRoboticsGame design
External game · account requiredCURIOSITY, MEET POSSIBILITY.
Build what
makes you
curious.
Your interests. A product prototype.
A new way to co-build with AI.
An education experience for high-school students.
MANTUO / OPEN LAB
○✳□
Make the invisible visible.CONCEPT DEMO
Drag to rotate · Change the energy01 — 03
26project groups sold
18student teaching sessions
3 + 1engineers + research mentor
5countries represented by students
01 / THE POSSIBILITIES
Ideas you can
actually explore.
Science, data, robotics, everyday problems.
Different interests open different doors.
HarryPRODUCT PROTOTYPE
A robotics arena you can enter from your browser.
3DRoboticsGame design
BrucePRODUCT PROTOTYPE
Change a parameter. See the physics change.
PhysicsSimulationLearning
LucyPRODUCT PROTOTYPE
Turn a content question into structured analysis.
MarketingDataResearch
MaddoxPRODUCT PROTOTYPE
Explore the systems behind investment research.
QuantDataSystems
HarryPRODUCT PROTOTYPE
Give a physical storage problem a digital interface.
ESP32HardwareIoT
5 projects to explore
A WINDOW INTO THEIR WORLD
A homepage.
A point of view.
Step beyond a single project.
Visit the student’s own space.
HARRYXIN↗
Harry XinRobotics, software & more ↗
RESEARCH
Hanchen ZhouEnter an animated world ↗
02 / THE EDUCATION EXPERIENCE
A prototype is the start.
Co-building is the skill.
Learn to notice a problem, work with AI,
make decisions and test what you build.
01 /
What if your interest could become a product?
You do not need to arrive with a complete idea. Bring a question, a frustration or a subject you love. We help turn it into a feasible prototype and a way of working with AI you can use again.
Notice problemsDirect & question AITest & improve
Take the next step →From “what if?” to “I built it.”
ONE PROJECT. YOUR OWN PATH.- 01
Discover
Start with your interests and a problem worth exploring.
- 02
Define
Discuss the idea. Engineers assess feasibility. Confirm a clear project plan.
- 03
Co-build
Work with AI and mentors across 18 teaching sessions. Test, revise, understand.
- 04
Share
An open-source repository, a working presentation and a record of learning.
03 / THE PEOPLE & THE FOUNDATION
Curiosity needs
a strong foundation.
Engineering experience. Research guidance.
Years of working with ambitious students.
3+1
Engineers + research mentor
Three engineers with frontline experience at major technology companies, supported by one research mentor.
Knowledge that
agents can reach.
IBIGCSEAPA-Level
Concepts
Questions
Teaching resources
Questions
Teaching resources
RAG
Database
Database
Learning agents
Project support
Project support
A growing, structured knowledge base drawn from international high-school curricula. Learning agents and standalone database products are future directions under exploration.
Continuously being builtMeet our research & engineering advisors
Zhang Xiao
Computer vision & deep learningCUHK MMLab PhD · University of Pennsylvania postdoctoral researcher
Meng Yiren
Research & science educationTongji Computer Science · Fudan Software Engineering
Liu Yanbo
Automation & embedded engineeringShanghai Jiao Tong University · Huawei Cloud MVP
Rong Xuning
Software architecture & deliveryBilibili gaming platform · Dewu trading-system architecture
MANTUO / 让成长自然发生From New York.
From New York.
Built across borders.
2017New York2018Shanghai
FOUNDER
Allen LuMantuo CEO
2013–2017 · PwC New York
Investment Banking & Capital Markets
Associate → Assistant Manager
In 2017, exposure to a J.P. Morgan financial-contract AI review project offered an early view of AI in professional work.
GROWTH, WITH A DIRECTIONLearn the skill.
Learn the skill.
Show the substance.
Nearly a decade of undergraduate-admissions experience helps us connect a student’s interests, project evidence and application narrative.
HarvardYalePrincetonStanfordMIT
Undergraduate admissions achieved across Mantuo’s study-abroad business over the years. These are cumulative company outcomes, not admissions attributed to Lab projects.
Project portfolioResearch & maker supplementsSuitable competitions
See MIT’s maker & research portfolio requirementsWe help identify suitable opportunities and document student work. Competition eligibility and AI rules are assessed individually; awards and admissions are not guaranteed.
Conrad Challenge
Explore a prototype’s purpose, users and presentation.
Regeneron ISEF
A research direction assessed against affiliated-fair eligibility and student-independence rules.
HiMCM
Pre-contest skill building. Students work independently during the contest and disclose AI use.
How rare are these admissions?
For the 2025 intake, incomplete industry tallies of known offers to students at mainland Chinese high schools reported roughly: Harvard 8, Yale 14, Princeton 9, Stanford 30 and MIT 5. These are contextual industry figures, not Mantuo’s admissions counts or fixed annual university quotas.
2025 industry report · scope and figures04 / YOUR NEXT CHAPTER
Bring an interest.
Let’s find the possibility.
For students in China, the United States,
South Korea, the United Kingdom and Canada.
SOFTWARE PROTOTYPE
From an idea
to an interactive product.
RMB52,000
Standard project
- 18 student teaching sessions
- 2 internal engineering R&D units
- Open-source code + deployed website
CHIP-BASED HARDWARE PROTOTYPE
Connect your idea
to the physical world.
RMB68,000
Arduino, ESP32 or similar chips / boards
- 18 student teaching sessions
- 2 internal engineering R&D units
- Open-source code + video / photo presentation
Good questions.
Clear answers.
Do I need to know how to code?
Start with an interest and the willingness to learn. We assess the idea, your starting point and a realistic project scope before the project plan is confirmed.
What does the student actually do?
Identify the problem, make design choices, work with AI, test the prototype and explain the result. Mentors guide and engineers support delivery; the student’s contribution should be visible.
How are lessons scheduled?
The project plan sets milestones and a schedule for 18 student teaching sessions. Two additional internal R&D units support engineering and deployment; they are not student lessons.
Is the final result a finished commercial product?
The target is a working product prototype with source code and an accessible presentation. Its specific functions and acceptance criteria are defined in the project plan.
How does it connect to university applications?
The process can create meaningful evidence of curiosity, decisions and technical contribution. Where appropriate, it may support portfolios, research supplements or competition preparation. Each route has its own requirements.
Can I explore without an account?
Yes. This showcase is public. Student websites open separately and may have their own access requirements; DriverLab currently needs a game account.