Inspiration

I study Petroleum & Gas Engineering at the University of Lagos, and I'm also a freelance developer. That means I work during the day and study at night.

The night before my hardest first-year exam, engineering mathematics, I left my hostel at 11:45 PM to study at the Faculty of Engineering, because my hostel locks its doors at midnight. I got there and there was no power. No light. Nothing. I couldn't study.

By the time I gave up and walked back, my hostel was already locked. I slept outside that night.

Lying there, one thought wouldn't leave me: what if I had known there was no power at the faculty before I ever left? In Nigeria, electricity is unpredictable everywhere, and nobody has real information about it. So I built the thing I wished existed that night. That's NEPAWatch.

What it does

NEPAWatch turns Nigeria's power grid into real-time, usable intelligence. It fuses community outage reports, weather, and other live signals, and an AI layer uses them to predict:

whether an area has power right now
how long the current supply is likely to last
when power is likely to be restored
how reliable a given area or report is, through trust scores

Underneath the predictions, NEPAWatch is quietly building something bigger: ground-truth data on Nigeria's electricity that doesn't exist anywhere today, not at NERC, not at the DISCOs. The predictions are the first product. The dataset is the real foundation.

How I built it

NEPAWatch is a full-stack web application. I built the frontend, backend, data pipeline and infrastructure myself.

Frontend & app: modern web stack, mobile-first for Nigerian users on mid-range phones and 4G
Backend & data: a live database of user-submitted grid reports, structured so every report becomes a data point the platform owns
AI layer: density-clustering and multi-source fusion (community reports, weather, live signals) to verify reports and generate predictions and trust scores
Infrastructure: I run the hosting, domains, SSL, backups and uptime myself

Traction

Launched with zero naira spent on ads:

1,500+ users across all 36 states of Nigeria
200 users and a 150-person community within the first two days
ITMO UNILAG sponsorship secured on day two, supporting IP protection
Selected to represent Nigeria's power sector at the UNDP innovation board meeting
Featured as an innovator at the AI UniPod launch at UNILAG
Covered by PurpleCom and invited to speak on Trade FM

Challenges I faced

The hardest part was trust. Crowdsourced data is only as good as its reliability, so I had to build a system that could tell a real outage report from a wrong or malicious one — which is where the AI trust-scoring came from.

The second challenge was doing it solo while studying full-time. NEPAWatch was built in the same night hours that caused the problem in the first place.

What's next

I'm building out the API infrastructure so other platforms can plug into NEPAWatch's data, and adding a payment layer for premium features and API access. The long-term goal: NEPAWatch becomes the data layer every energy company, DISCO, and regulator in Nigeria eventually builds on.