MarketBridge

Inspiration

In many Nigerian markets, vendors rely heavily on guesswork to make key decisions—what to stock, how much to charge, and when to restock. Growing up around these informal markets, I saw firsthand how even the most hardworking sellers often struggled with inconsistent income, avoidable losses, and zero access to loans or digital tools.

This project was born from a simple question:
"What if local sellers had a digital assistant that could help them make smarter business decisions every day?"

What it does

MarketBridge is a mobile-first AI-powered app that helps informal market vendors:

• Track daily sales and profit.
• Get smart restock suggestions.
• Receive price optimization tips.
• Forecast demand using local trends.
• Build credit profiles based on behavior data.

It's designed to feel simple, even for first-time smartphone users, while packing the power of machine learning under the hood.

How we built it

The stack includes:

• Frontend: Bolt (Flutter-powered builder)
• Backend: Node.js + Express
• Database: Firebase Firestore (for real-time updates)
• ML Engine: Python (Prophet for forecasting, Scikit-learn for clustering and scoring)
• Design: Figma + Bolt UI components
• Hosting: Firebase Functions

Challenges we ran into

• Data Scarcity: We had no access to clean historical sales data, so we created synthetic datasets and built models that learn as users interact with the app.
• Design Simplicity: Designing for non-tech-savvy users meant stripping the UI down to only what’s essential—no clutter, no jargon.
• Trust: Getting informal vendors to trust a “smart” app was tough. We designed onboarding with clarity and privacy in mind.
• Connectivity: We optimized the app to work offline and sync when back online.

Accomplishments that we're proud of

• Built and deployed a working prototype using Bolt in record time.
• Developed a lightweight but powerful ML engine for local trend forecasting.
• Created a user experience that's both accessible and empowering for informal vendors.
• Got strong feedback from initial test users in local Nigerian markets.

What we learned

• ML doesn’t have to be complex to be useful—it just needs to solve a real problem.
• Building for emerging markets means balancing ambition with simplicity.
• Tools like Bolt.new drastically speed up development when you want to move fast and test real ideas.
• Real impact starts with understanding your users deeply.

What's next for MarketBridge

• Partner with POS systems (like Moniepoint, Opay) to automate transaction tracking.
• Add voice input in local languages for accessibility.
• Expand the ML model to include seasonal trend detection.
• Launch pilot programs with local sellers and microfinance banks.
• Add a rewards system for consistent users (like vendor loyalty programs).

Built With

• bolt
• express.js
• figma
• firebase
• node.js
• python