We bring the computer. You bring a router and a power outlet.
Umanah AI Lab puts a full AI tutor, coding lab and learning library inside your school, free of monthly or yearly fees. Students log in from the devices they already have and learn for as long as they want, even with the internet unplugged. As a nonprofit, we supply the box.
Prototype #001 is live. The demo uses made-up students and sample data.
A real screenshot of the running demo. Sample data.
The access gap
AI for learning shouldn't depend on a connection, a vendor, or a monthly bill.
Many schools have unreliable internet. Student conversations leave the building and sit on someone else's servers. The rules change when a provider changes its terms. And a subscription for every student adds up fast.
Bring the intelligence to the classroom instead of sending the classroom to the cloud.
What it is
AI, curriculum, coding, and exploration in one place.
Students open a browser and sign in. Everything runs on one box at the school. Five actions cover the whole experience.
Ask
A tutor that gives help grounded in the student's lessons, and hints (not answers) in homework.
Learn
Courses, lessons, assignments, and a Revision page with weak topics, practice questions, and flashcards.
Build
Write and run Python in the browser, with an AI helper that explains errors and gives hints.
Explore
Search an offline library. It works with the internet unplugged.
Progress
XP, levels, badges, streaks, and projects: growth and accomplishment, not public ranking.
Inside the product
Real screens from the running prototype.
No mock-ups. These are screenshots of the demo you can open right now. All names, chats, and numbers are sample data.
Ask. Answers are formatted, cited from the lesson, and can be tried in Build.Revision. A ten-minute sprint built from what the student found hard.Build. Code runs in the student's browser. The helper explains errors.Explore. An offline encyclopedia with search.Learn. Courses, assignments, and revision in one place.Teacher Center. Class progress and students who may need help.
How it works
One box. Many devices. No cloud in the loop.
1
Plug it in at school
The appliance joins the school network and serves the whole class.
2
Open a browser
Students and teachers use the laptops, tablets, and phones they already have.
3
Learn, locally
The tutor, lessons, and library all run on the box. With the internet unplugged, learning continues.
Safety & privacy
Built for children, with the school in charge.
Student conversations stay at the school. They are not sent to Umanah.
Homework rules are enforced by the software, not just requested in a prompt. In homework the tutor gives hints. In assessments it stays out.
Every message is screened before and after the tutor answers.
Safety alerts go to the school. Messages that suggest a student may be in danger reach the school's designated safety contacts, never Umanah.
Teachers see what matters. Chats linked to an assignment are visible to the class teacher. Students can flag a tutor answer that seems wrong.
The safety rules are still being reviewed. Before any classroom pilot they are reviewed by counselors and safeguarding specialists.
Structured learning
More than a chatbot.
The tutor sits inside a learning loop: lesson, practice, tutor help, exercise, project, achievement. Numbers are computed by code, not guessed by the model.
Lesson
Practice
AI tutor
Exercise
Project
Achievement
Teacher Center
Teachers stay in control.
Add students one by one or from a spreadsheet, print login cards, set how much help the tutor may give on each assignment, grade work with a score and comment, and see which students may need help.
Classes, rosters, and login cards
Assignments with a hint ladder for every question
Grading with feedback that students are notified about
Flagged tutor answers and students who may need help
Assessments, class resources, and an AI lesson builder are coming.
Same platform, different sizes
Deployment configurations
One platform, sized to the setting. Capacity figures will be published once they are measured on the final hardware. Today the prototype runs on a single development server.
AI Lab Mini
A compact appliance that serves a classroom, library, or small community space.
A small group of students at a time
Good for pilots and introductory use
AI Lab School
A shared deployment for an entire school or several classrooms.
More capacity for more concurrent users
Shared across teachers and students
AI Lab Research
Higher-performance deployment for advanced use and research.
More powerful hardware and storage
Larger models and datasets
Runs locally. Your data stays on your premises.
Shared by many students. Designed for classrooms and communities.
Works with limited connectivity. Core learning needs no internet.
Built to be maintained. Health checks and updates are on the roadmap.
Open infrastructure
Mature open-source technology, unified.
Umanah LabOS brings proven open-source software together into one simple experience for schools, so no one has to assemble it themselves.
Includes open educational resources from leading open-content projects. The offline library in the demo uses Simple English Wikipedia, shared under the Creative Commons Attribution-ShareAlike licence.
Open content is credited where it is shown. Endorsements are never implied.
Pilot program
Bring local AI to the places that need it.
We are looking for schools, libraries, nonprofits, community organizations, technology and hardware partners, and philanthropy who want to help shape the first pilots.
Tell us a little about your setting. We read every request.
No pilot has started yet, and no partner relationships are implied by this page.
One box at the school. Many devices. No cloud in the loop.
