UNISTORE SEARCH

In a country where conventional ecommerce is seeing a drastic reduction in adoption rate, unistore search is an AI powered platform offering the next solution.

Comment 1

Inspiration

You know, the journey to UniStore Search wasn't a sudden flash of genius; it was a gradual unfolding, a story rooted in years of working in e-commerce here in Nigeria with my co-founder and friends. We always had this big vision: to create an e-commerce platform specifically for universities, starting right here at home.

We launched so many MVPs over the years, constantly tweaking and learning. Then, this past April, we finally launched unistore.ng. We were thrilled! We quickly got over 700 products uploaded and built a vibrant community of over 2,000 members across our WhatsApp groups. Although we've just been operating majorly in only 2 universities, Bingham and Veritas university here in Nigeria.

But something unexpected started happening. We thought users would browse the site, find what they needed, and then chat with merchants, just like conventional e-commerce works. Instead, time and time again, the majority of product requests came in a different way: users would just send us a direct text asking, "Do you know any seller who offers [specific product]?"

It was a lightbulb moment. This could be a major reason why e-commerce user adoption rates in Nigeria seemed to be declining. It wasn't just about a website; it was about a culture shift. While some people certainly enjoy Browse through a catalog, a huge portion of users, we realized, prefer to make an on-the-spot request for a specific product they need and connect directly with a seller, rather than sifting through endless products with an e-commerce company acting as a middleman.

This fundamental insight, this user behavior, is what truly birthed the idea of UniStore Search.

The Larger Picture: Why UniStore Search Matters for Nigerian SMEs The story of UniStore Search isn't just about a new feature; it's about addressing the deep-seated challenges faced by Nigerian Small and Medium-sized Enterprises (SMEs), which are the backbone of our economy. These businesses make up a staggering 96% of all businesses, contribute 48% to our national GDP, and account for 84% of our employment. Yet, they face brutal realities:

Alarming Failure Rates: Over 50% of SMEs fail in their first year, and more than 95% don't make it past their fifth anniversary. This isn't just a statistic; it's countless dreams shattered.

The Struggle for Customers: A primary killer is the lack of market demand, with over 40% of failures directly attributed to an inability to find customers. Imagine having a great product but no one knows you exist, trapped among the approximately 39,654,385 MSMEs operating in Nigeria as of December 2020.

UniStore Search directly tackles these issues head-on:

Solving the "Finding Customers" Problem: Instead of merchants waiting to be discovered, UniStore Search empowers users to actively seek products. Our Google Gemini AI uses natural language processing to understand specific product requests, then intelligently matches these immediate needs with relevant sellers. This ensures targeted demand meets supply directly, vastly improving an SME's chances of making a sale and surviving.

Adapting to User Culture: Recognizing that a large segment of Nigerian users prefers direct, on-demand connections over Browse catalogs, UniStore Search eliminates friction. It bridges this cultural gap by facilitating direct communication via WhatsApp integration, making transactions feel personal and immediate.

Fairness and Visibility: Our sophisticated matching algorithm prioritizes sellers with high ratings but also uses a recency-based visibility system. This means smaller or newer businesses aren't lost in the crowd; they get a fair chance to be seen and connect with potential buyers, breaking down barriers to entry.

Building Trust and Insights: Through our built-in rating and review system, sellers can quickly build a trusted reputation. For us, the administrative dashboard provides crucial analytics on user requests and seller performance, giving us valuable data to continuously refine the marketplace and offer better insights to our merchant community.

Right now, UniStore Search thrives by connecting sellers within universities to student buyers. But our ambition stretches far beyond campus gates. The ultimate vision is to expand nationwide, leveraging geolocation proximity-based matching. Imagine being able to find any product or service you need, from any local seller, just based on your location. This will unlock immense potential for millions more SMEs across Nigeria, turning every corner of the country into a vibrant, accessible marketplace.

What it does

Unistore Search is an AI powered online platform designed to connect university students with local sellers, built on a robust technical foundation.

For Students: When you submit a product request, Unistore Search employs AI-powered natural language processing to interpret your needs and identify relevant product categories. A sophisticated matching algorithm then connects your request with available sellers, prioritizing those with high average ratings and ensuring fair exposure through a recency-based visibility system. You can directly engage with sellers via WhatsApp link integration. Post-interaction, a built-in rating and review system allows you to provide feedback, which dynamically updates seller profiles. All your search requests and interactions are logged for historical reference.

For Admins: Administrators manage the platform through a dedicated dashboard, performing CRUD operations (Create, Read, Update, Delete) on seller profiles, product categories, and university settings. The dashboard provides analytics on user requests, visitor engagement, and seller performance, leveraging data from the rating system and contact interactions. A critical feature is the management of seller billing statuses, which automatically determines their active/inactive state and influences their visibility in search results. The system is powered by a Supabase backend for data management and Google Gemini AI for intelligent matching.

How we built it

  • Told bolt to create the home page search UI.
  • Then I told bolt to create a way to add sellers using a list of seller information i collected using google forms.
  • Next i told bolt to use gemini AI model to dynamically generate seller product categories based of the seller description and store it to the seller profile and a category catalogue.
  • Next I told bolt to create product categories based of users request made on the home page and match it to the list of categories catalogue then match with relevant sellers in the database.
  • Overtime added more prompts to refine the matching algorithm, included user reviews and included analytics to track the entire happenings on the site.

Challenges we ran into

  • The matching algorithm was the biggest challenge for me.

What's next for UNISTORE SEARCH

We're focused on supercharging UniStore Search's growth and impact. Our immediate goal is to deepen adoption within our two primary university campuses while simultaneously expanding to other universities across Nigeria.

Looking ahead, we're building the future of commerce by training our own custom AI model using our data. This will enable us to deliver extremely effective and personalized seller recommendations based on users' requests for products and services, making user discovery incredibly efficient. Furthermore, we plan to integrate our seller response API with external sites. This will allow other platforms to embed our direct seller-matching capabilities, laying the groundwork for the next generation of commerce infrastructure in Nigeria, Africa, and globally.

Built With

Share this project:

Updates