Inspiration
At the start of my first semester in college, I realized I often forgot what I had learned shortly after class. I was able to pass tests because I could force myself to focus and cram — but it wasn’t sustainable. When my results came out, I nearly failed one subject I disliked. That moment hit hard. I told my mentor, and he skipped over my good grades to point out the D. That’s when I realized: in my family, excellence is the bare minimum — and I had to raise my standard.
So I decided to build a system for myself and others who struggle not with motivation, but with knowing how to learn. I wanted a tool that gives students clarity — and during my research, I found the Yoruba word ELARO, meaning “clarity of thought.” It felt perfect.
In January 2025, I wanted to start an AI Automation Agency, but my CS degree was eating up all my time. I joked to a friend, “I need to hire an academic assistant.” That’s when it clicked: maybe I could build one. I explored study apps and realized they all give you tools but no guidance. It's like giving someone who can't drive a Lambo and expecting them to win races. I saw a gap.
When my dev said the mobile app would take a couple of months, I didn’t wait. I stumbled upon Bolt through the Vibe coding community and within 6 hours, I had a working prototype. I was hooked. With ChatGPT as my guide and Bolt as my engine, building ELARO felt less like work and more like playing a video game. And I just kept going.
What it does
ELARO v0.5 helps students discover how they learn best and gives them actionable tips on how to study smarter — not harder.
The system includes: A smart quiz that blends multiple learning science models (VARK, Kolb, Felder-Solomon) A breakdown of the student’s learning profile Tailored tips based on their learning style Even though our full AI-powered system isn’t currently online due to token exhaustion, students can still complete the quiz and receive useful feedback. We’ve temporarily shifted to a fallback system that ensures users can still discover their learning style and get guidance — the core promise remains intact.
How We Built It
We built everything with Bolt.new, powered by AI tools like ChatGPT, Claude, and Manus AI. Bolt handled the entire UI and backend logic — both the quiz system and result delivery. We used Claude for intelligent question generation and personalization. ChatGPT served as my technical cofounder — helping write logic, fix bugs, and create content like our legal docs. While the ELARO Map and advanced result generation features were working during development, they are currently offline due to API token limitations. We plan to restore them after the hackathon. Despite these limits, users can still take the quiz and receive a learning breakdown that makes them feel understood and equipped.
Challenges We Ran Into
Token exhaustion meant our Claude and ChatGPT-powered features had to be paused right before the deadline. We had to quickly build a fallback system so users could still get value from the quiz — a decision that forced us to be scrappy and prioritize user trust over polish. Technical complexity was also a barrier, but AI helped me overcome it. I learned how to reduce prompt complexity and focus on what matters most. I had zero coding experience when I started. Bolt, AI, and the desire to help students kept me going.
Accomplishments We’re Proud Of
Built a functional, personalized learning style quiz from scratch with no tech background. Enabled over 20 students from different universities to discover how they learn best — with great feedback. One tester said: “Now I finally understand why my current strategy doesn’t work.” Created a full product vision — from intake to personalized plan — even if not all features are currently active. Documented the journey publicly, inspiring other students and first-time builders across Ghana.
What We Learned
You don’t need to know how to code to build a meaningful product — but you do need to know how to explain your vision to AI. Bolt taught me the value of simplifying features, building fast, and testing even faster. I learned how to design with constraints — a skill that’s even more important when you don’t have budget or experience.
What’s Next for ELARO
We’ll restore the ELARO Map and advanced AI-generated results after upgrading our token capacity. Our mobile app (currently in development) will evolve ELARO into a full academic co-pilot — helping users plan, revise, and stay on top of their coursework. We're starting with students at KNUST, but we’re already reaching other universities across Ghana. I’m documenting this journey under “Building a Tech Startup in Ghana with No Tech Experience,” and sharing it to inspire the next wave of student builders. No matter how far we scale, ELARO will always be about clarity, personalization, and empowering students to succeed their way — not someone else’s.
Built With
- bolt.new
- claude
- supabase
Log in or sign up for Devpost to join the conversation.