Community-Contextualized Learning · South/Southeast Asia

Maati

Learning that starts where people are

Maati is a community knowledge platform built around what people already know — not what an imported curriculum assumes they should learn. Designed for mixed-age learners in low-resource settings, offline-first, in local languages.

See the approach → Research design

"The issue is not that communities lack knowledge — it's that the knowledge they have is invisible to the systems designed to help them."

— Adapted from Moll, Amanti, Neff & González (1992), Funds of Knowledge

739M
Adults globally lacking basic literacy skills
UNESCO, 2024
77%
Concentrated in Sub-Saharan Africa & South Asia
UNESCO, 2024
70%
Of digital skills demand in Africa is foundational
ADB / UNIDO, 2022
80–90%
EdTech programs in Asia estimated to fail
ISTE field report

The problem

The gap isn't access.
It's relevance.

Most edtech built for the Global South is designed elsewhere and shipped in. It imports not just content but assumptions — about literacy levels, about learning contexts, about what counts as useful knowledge. The result is predictable: communities get tools they cannot use, for problems they were not asked about.

2.7B
People still offline
The digital divide affects over a third of the global population — concentrated in low-income countries where foundational digital skills matter most for economic participation.
ITU, 2023
62%
Children in SE Asia & Sub-Saharan Africa never assessed
Over half of children in these regions have had their reading and learning skills never measured since 2015 — a fundamental data blind spot that persists into adult education.
UNESCO / WEF, 2024
1%
Platform usage in Malaysia's flagship program
Malaysia's 1Bestari Net program — broadband for all schools, virtual learning for students and teachers — achieved under 1% active use and was discontinued. A pattern, not an exception.
ISTE, 2015

"Experts estimated 80 to 90% of edtech programs in Asia were failures, with most schools unable or unwilling to use the technology effectively."

— ISTE international consultant field report, Asia

The failure pattern is consistent across contexts: content that doesn't reflect local language or culture, teacher training absent or inadequate, no alignment with what learners actually do. Research from Ghana found that even the accent in a reading program's narration caused confusion — because it was American English in a Ghanaian classroom.

Korsah et al. (2010), via Wiley Review of Education

The approach

Built with, not for.

Maati is grounded in funds of knowledge pedagogy — the idea that every community has rich, practical knowledge that formal systems routinely ignore. Instead of importing a curriculum, we build around what's already there: how people price goods, manage savings, grow food, and use their phones.

1
Community listening before building
Month 1 is entirely dedicated to structured listening — no content created, no assumptions validated. We ask what people already know, what problems they're solving, and what a "useful lesson" would look like in their context.
2
Offline-first, mobile-first design
Every lesson works without an internet connection. Built on Kolibri or as a Progressive Web App, content is downloaded once and runs on any Android device — the technology most learners already own.
3
Facilitators, not instructors
Delivery is through trusted community members — a teacher, a youth leader, a savings group coordinator — who co-design the lessons and run the sessions. The platform supports them; it doesn't replace them.
4
Modular and locally selectable
No community takes all four modules. Each stands alone so communities choose what's relevant. A fishing village and a market town have different needs — the platform shouldn't pretend otherwise.
MVP · In development
Digital Literacy for Real Life
WhatsApp for business, scam awareness, free tools like Google Sheets and Canva. Starting with the phone people already have.
6 lessons · Pilot ready Month 3
Module 2 · Planned
Small Business & Entrepreneurship
Pricing, tracking sales, mobile money, and building a customer base within existing market structures.
Expanding Month 5
Module 3 · Planned
Personal & Household Finance
Budgeting, savings groups, avoiding predatory lending — built around the financial tools the community already uses.
Post-pilot
Module 4 · Planned
Agriculture & Food Systems
Soil health, seasonal planning, cooperative selling, and reducing post-harvest loss.
Post-pilot

Research design

A question, not a claim.

Research question
When digital learning content is co-designed with community members around their existing economic practices, does learner engagement and self-reported knowledge application differ from engagement with externally-designed curricula — among mixed-age, limited-schooling adult learners?

This is not a controlled trial. It's a design-based inquiry — documenting what happens when a community is treated as the expert, not the recipient. The goal is to generate questions worth studying further, not to prove a universal answer.

  • 📋
    Attendance & engagement tracking Session-by-session attendance patterns and facilitator observation notes using a simple structured rubric.
  • 🎙
    Voice-note knowledge application interviews 3-question post-lesson voice note: "What did you do differently this week because of what you learned?" No literacy required.
  • 📓
    Co-design decision log Every time a learner or facilitator says "this doesn't fit how we do things here" and the content changes — documented. This is the evidence of the funds of knowledge approach in action.

Theoretical grounding

Funds of Knowledge
Moll, Amanti, Neff & González (1992)
Core framework
Generative Themes
Paulo Freire, Pedagogy of the Oppressed
Philosophical anchor
Capacity to Aspire
Arjun Appadurai (2004)
Equity framing
Design-Based Research
Brown (1992), Collins (1992)
Methodology
SDG 4.4 — Digital Skills
UNESCO / UN, ongoing
Policy context
Applying to
Harvard HGSE — LDIT Program Learning Design, Innovation & Technology. This project addresses their core question: what does equitable, technology-mediated learning actually look like outside OECD contexts?
Columbia Teachers College — TML Program Technology, Media & Learning. The scam awareness and digital footprint modules speak directly to their critical media literacy thread.

6-Month plan

The project as it unfolds.

Month 1
Community listening
Structured conversations before any content is built. Understanding needs, existing knowledge, and what "useful" looks like locally.
Now
Month 2
Build MVP module
4–6 short lessons on digital literacy. Visual-heavy, translated, designed for no-literacy-required delivery.
Upcoming
Month 3
Pilot with 10–15 people
Delivered through a community facilitator. Observation notes, voice interviews, co-design log running.
Upcoming
Month 4
Iterate on feedback
What didn't fit? What surprised us? Revise content, revise the model. Document every design decision.
Upcoming
Month 5
Expand to Module 2
Small business & entrepreneurship. Same co-design process. Second facilitator if the community has grown.
Upcoming
Month 6
Document & publish
Case study brief on Medium or a simple site. What was learned — honestly, including what didn't work.
Upcoming

Results — Pilot pending

This section is intentionally empty.

The pilot hasn't run yet. What you see here are the metrics this project will track — and a placeholder for what the community teaches us. Results will appear here from Month 3 onward.

—
Pilot attendance rate
Sessions attended vs. sessions offered across the 10–15 person pilot cohort. Primary engagement signal.
Available Month 3
—
Self-reported application
Percentage of learners reporting at least one behavior change in the week following a lesson, via voice note interview.
Available Month 3
—
Co-design changes logged
Number of content decisions changed based on direct community feedback. This is the funds of knowledge principle made visible.
Running from Month 2
🖥
Prototype in development The first two lessons of the Digital Literacy module are being built in Months 1–2. A live demo will be linked here once available for facilitator review. The prototype is designed to run on a low-end Android device with no internet connection.

About this project

Who built this — and why it matters to say so.

Maati was designed by someone with a computer science and UX background, family ties to South/Southeast Asia, and a growing frustration with how edtech treats the Global South as a deployment target rather than a design partner.

The theoretical grounding came before the prototype. The community contact is still being built. That sequencing — research first, humility first — is intentional.

A note on proximity
I have family ties to the pilot region but not lived experience in low-resource community contexts. This is why the co-design structure exists — not as a methodology checkbox, but as a genuine design constraint. I don't know what this community needs. That's the whole point of Month 1.
🎓
Background
Undergraduate in computer science. Experience in product and UX. Now focused on where design thinking meets education equity in low-resource contexts.
📚
Grounded in the literature
Built on Moll & González's funds of knowledge framework, Freire's generative themes, and design-based research methodology. The theory preceded the prototype.
🌐
Looking for
A community facilitator — someone already trusted in the pilot community who wants to co-design, not just deliver. Not a gig; a co-author.
✉️
Get in touch
If you work in community education, EdTech for the Global South, or international development — or if you know someone who should be part of this — reach out.