Inspiration
In Burundi and most of the world, students don't have internet or mobile phones but almost everyone has a basic mobile phone. We asked ourselves: "What if learning was as simple as sending a text?" And so SMS Tutor AI was born.
What it does
SMS Tutor AI transforms ordinary phones into personal tutors, so that, students can send invitations like, “Explain fractions" via SMS, and get back clear, precise, step by step explanations written in either English, French or Kirundi. Teachers can send quizzes via SMS, so that students can respond with answers via SMS. Parents receive brief updates on student progress simply through SMS . No apps, no internet, no barriers, just learning using SMS.
How we built it
We integrated AI with the most straightforward technology: text messaging. A Python backend connects all SMS requests to the AI model. An ESP32 with a SIM800L GSM module sends and receives SMS messages directly, without any use of the internet. A small database stores student records and quiz results. Multilingual support ensures community inclusivity.
Challenges we ran into
Establishing reliable two-way SMS communications with SIM800L and ESP32 hardware. Decrease downtime so answers feel instantaneous. Simplifying explanations to the level of younger students. Designing for classrooms that have very limited resources.
Accomplishments that we're proud of
Built a complete working prototype: a student sends an SMS → the AI reads it → the ESP32 returns a distinct answer. Shown that AI learning without internet through only GSM networks is possible. Built a solution based on real local requirements, but scalable around the world.
What we learned
On How to Combine AI models with Hardware (ESP32 + GSM) That ease of access is what matters most over complicated functionality Even simple tech like this , when used with Artificial Intelligence , can create very huge social impact.
What's next for SMS Tutor AI
Implement pilot programs in schools in Burundi to solicit feedback from both students and teachers. Scale to additional languages like Swahili and Kinyarwanda. Include voice/Speech-to-SMS function to support students who have difficulties with reading. Collaborate with NGOs and Ministries of Education to deploy SMS Tutor AI in classrooms across Africa and beyond.
Built With
- arduino
- esp32
- fast
- gsm
- llm
- natural-language-processing
- python


Log in or sign up for Devpost to join the conversation.