VerixamIR

The AI-Powered Automated Image Suite for Exam Registrations Centres.

Comment

Inspiration

Over the Years while working as a freelancer IT suppport i have seen parents, schoo administrators and Enrollment agents pleading with an official because there students registration was flagged for a "Bad Image." That mistake didn’t just mean a blurry photo; it meant a heavy penalty fee she couldn't afford and the risk of her son missing a year of school.

In Nigeria and across West Africa, we have over 20,000 exam centers, but the tools they use are manual and not customized to fit the purpose of schools. I realized that WAEC already captures perfect HD photos—why should a student have to pay, travel, and re-capture that same photo for NECO or NABTEB? I built VerixamIR to be the bridge that saves families money and keeps over 30,000 students in the classroom across all centres. It will also make it easy for freelance agents like myself who like to save every penny to avoid doing one work twice, and parents save over $100,000 in penalty fee

What it does

VerixamIR is a "smart assistant" for exam enrollment officers. Instead of wrestling with Photoshop or online croppers that ruin image quality, officers can process an entire school’s worth of students in minutes. It automatically snaps images to the exact pixel requirements of WAEC, NECO, and JAMB. Most importantly, it "talks" to existing WAEC databases, allowing centers to reuse high-quality, verified photos for other exams instantly. It’s about making enrollment move at the speed of thought, not the speed of paperwork.

How we built it

I chose Python because of its power to handle images delicately. -The Look: I used CustomTkinter to give it a modern, intuitive feel—something that feels like a 2026 app, not a Windows 95 utility.

-The Brain: The core engine uses a recursive logic I developed; it "tests" the image quality multiple times until it hits that tiny "sweet spot" (like 12KB) required by government servers.

-The Safety: I implemented multi-threading so the app stays responsive even when it's processing 1000 students at once.

Challenges we ran into

The biggest headache was the "KB Wall." Government servers in West Africa are incredibly strict—if an image is 16KB, it’s rejected. If it’s 8.9KB, it’s rejected. Coding a mathematical loop that could shrink a 5MB smartphone photo down to exactly 12KB without making the student look like a pixelated ghost was a true test of patience. I also spent many late nights reverse-engineering how legacy .cnt database files store student data to ensure our extraction was 100% accurate.

Accomplishments that we're proud of

I’m most proud of the Database Extraction feature. Seeing the software pull a student's name and photo directly from a WAEC file and turn it into a perfect NECO upload in two seconds felt like magic. We aren't just saving time; we are eliminating the "Identity Crisis" where students' names are spelled differently across different exam boards. I also tested with over 10 centres across Nigeria states and 1 from Togo which gave and got good feedback between as it worked seamlessly.

What we learned

This project taught me that innovation isn't always about building something brand new—it's about fixing what's broken. I learned more about JPEG compression than I ever thought I’d need to know, but more importantly, I learned that small technical solutions can have a massive social impact on education accessibility.

What's next for VerixamIR

The goal is to go completely "Hands-Free." I’m working on an AI Validator that will use facial recognition to tell an officer, "Hey, this student is looking away," or "The background isn't the right shade of red," before they even try to upload it. I want to expand this into a mobile app so enrollment agents can help students in the most remote villages, ensuring no child is left behind because of a technicality. I also wish to publish same Application to Microsoft and Play Stores for wider reach.

Built With

Share this project:

Updates