Building Leap From a Dorm Room
What it looks like to build an AI startup in Nigeria as a full-time student.
There's no romantic version of this story. Building Leap from a Nigerian university campus means loud generators, inconsistent Wi-Fi, and learning to architect distributed systems between lectures on business administration.
I'm Oluwanifemi Bello, and I founded Leap while enrolled full-time at Bells University of Technology in Ota, Nigeria.
What Leap Is
Leap is an AI-powered action engine. Not another chatbot. Most productivity tools optimize for input: they help you think, plan, organize. Leap optimizes for output. You describe what needs to happen. Leap makes it happen. The extraction engine at the core reads intent and produces structured, executable action across APIs, browsers, and workflows.
The Build Environment
I don't have a desk. I have a bed, a monitor, and a power bank I charge at the library when NEPA takes the light. The stack, Next.js, TypeScript, Python extraction pipelines, was chosen partly for performance and partly because I can run it locally without cloud uptime being a given.
What I've Learned
Shipping beats planning. Always. The Leap extraction engine has been rewritten three times. Each rewrite was faster and cleaner, because each version before it taught me what I was trying to build.
If you're a founder in a resource-constrained environment: your constraint is not unique. Your response to it can be.
— Oluwanifemi Bello (Nifemi Bello), Ota, Nigeria.