This article relates to Goal 4: Quality education, one of the 17 UN Sustainable Development Goals. Goal icon by the United Nations. Emberoot is not affiliated with, or endorsed by, the UN.
Education technology is often sold as the shortcut to SDG 4, the UN goal on quality education. The evidence is more modest. Devices handed to schools without training, content in the right language or time in the timetable tend to end up in cupboards. The tools that help are usually simpler: national platforms that put textbooks and lessons online, offline libraries for schools with no internet, and data that helps teachers teach children at their actual level. This piece looks at named examples and their limits.
Start with the gap EdTech has to cross
The UN’s Sustainable Development Goals Report 2026 says fewer than half of primary schools worldwide have computers, internet or facilities adapted for students with disabilities, and more than a fifth lack electricity, safe water or basic sanitation. In the least developed countries, about three quarters of primary schools lack computers and internet. Any education technology for SDG 4 has to work in those conditions, or it will widen the gap between schools that have and schools that do not.
National platforms: DIKSHA in India
DIKSHA (Digital Infrastructure for Knowledge Sharing) is India’s national platform for school education, run under the Ministry of Education with NCERT. It hosts digital textbooks, videos, courses for teachers and material for programmes such as NIPUN Bharat, the national mission on foundational literacy and numeracy. Many printed textbooks carry QR codes that open related lessons on a phone. The app lists more than 5 crore downloads.
The lesson for other countries is the link to the printed textbook. A teacher or parent who already has the book can reach extra material in one step, without searching an app store or a website.
Offline learning: Kolibri
Kolibri, from the non-profit Learning Equality, is a free, offline-first learning platform. A school or community centre can load lessons onto one local device and share them over a local network with no internet connection. Learning Equality says it has been used in more than 220 countries and territories, with library content in over 173 languages. Offline-first design is the right default for most of the schools SDG 4 is worried about.
Arabic-language content: Madrasa
Madrasa is a free online platform of science and maths video lessons in Arabic, covering kindergarten to grade 12, run by the Mohammed bin Rashid Al Maktoum Global Initiatives foundation in the UAE. Language is one of the biggest barriers in online learning, since much free content is in English. Platforms that teach in the language children speak at home and at school reach far more of them.
Data for teaching at the right level
Some of the strongest evidence in education comes from an approach that uses data rather than devices. Teaching at the Right Level, created by the Indian NGO Pratham, assesses children with a simple test and groups them by learning level instead of age or grade. J-PAL reports that randomised evaluations in India found large learning gains when it was implemented well, and that the approach has reached over 60 million students across India and Africa. TaRL Africa has supported implementation in 20 countries.
J-PAL also notes that training or materials alone did not improve learning. Gains came from clear goals, regular use of data, on-site mentoring and peer support for teachers. Software that helps teachers record levels and form groups can make this easier, but the method is the active ingredient.
Where AI fits, carefully
AI tutors and lesson planners are spreading fast. Used well, they can help teachers draft questions, adapt texts to different reading levels or give students practice with feedback. Used badly, they hand children confident wrong answers and collect their data. The safer starting point is teacher-facing tools, where a trained adult reviews output before it reaches a child, with clear rules on age limits and data retention.
The limits
- Devices without teachers: technology works through teachers. Training and time to use it matter more than hardware.
- Connectivity and power: assume no reliable internet, and design for offline use and low-cost phones.
- Language: content in a language children do not speak well teaches little.
- Privacy: children’s data needs strict protection, with consent from parents and schools.
- Screen time: digital practice should add to, not crowd out, reading, writing and talking.
Before building a learning app, spend a day in a classroom like the one you are designing for. Count the working devices, sockets and minutes the teacher really has.
Building learning tools responsibly
If you are planning a learning platform, school system or tutoring app, start from the conditions your users face. Our SaaS and app development work begins with a free 30-minute call and a written scope agreed before any build. For the targets and data behind this goal, see our Goal 4 page and the wider Global Goals hub.
Frequently asked questions
Does education technology improve learning?
It can, when it supports good teaching and fits local conditions. Tools that work offline, use local languages and help teachers match lessons to children’s levels have the best record. Devices handed out without teacher training, suitable content or time in the timetable usually show little effect on learning.
What is DIKSHA?
DIKSHA, short for Digital Infrastructure for Knowledge Sharing, is India’s national digital platform for school education, run under the Ministry of Education with NCERT. It offers digital textbooks, videos, teacher courses and material for national programmes such as NIPUN Bharat. QR codes printed in many textbooks link to related lessons on a phone.
What is Teaching at the Right Level?
Teaching at the Right Level is an approach created by the Indian NGO Pratham. Teachers test children’s reading and maths with a quick assessment, group them by level rather than grade, and teach activities matched to each group. Randomised evaluations reported by J-PAL found large gains in India, and the approach has spread to many African countries.