Mali-Agent

Inspiration

The inspiration for this project came from the frustration of modern online shopping. We’ve all faced that frustration at once. Too many choices, too much time wasted, and too little efficiency. Consumers struggle to find exactly what they need, while sellers get buried under endless listings, making it difficult to reach the right buyers. We thought of a smarter way to shop. One where users simply type what they want, and AI instantly connects them to the best options.

What it does

It allows users to search for a specific product, order it for them and handle the payment process with the user using only their words as if talking to their assistant.

How we built it

For this testing phase, we collected various local Restaurants with their different menus and prices into one Data set. We then used an intuitive chatbot combined with Open AI model GPT4o to search for any entered menu by the user and make suggestions to them of different restaurants that have the menu they crave. From the top-rated restaurants to the least rated. With just their words, the user is able to tell to the AI-Agent look for a specific menu and to pass the command for them. At the end they will just enter their phone number for payment and have to confirm the transaction by entering their mobile money account PIN.

Challenges we ran into

The biggest challenges we faced was the payment calls and writing an efficient system prompt for the AI model to fully understand the user’s requests.

Accomplishments that we're proud of

We're proud of being the ones to tackle the issue of frustration everyone encounters when dealing with delivery apps. Saving people's time makes feel that we have a great super power. And it's said that "a great super power implies great responsibilities".

What we learned

In few words, building this project taught us the power of AI in simplifying complex problems. We explored chatbot technology and recommendation algorithms to make searching for products as effortless as messaging a friend. Additionally, we learned how crucial user experience is in reducing decision fatigue.

What's next for Mali-Agent

With Mali-Agent, we aim high. These are our next steps:

Expanding to more African cities while adding new local languages and speech to text way of searching.
Building a delivery network to connect rural areas with urban food options
Adding AI-powered menu recommendations based on user preferences and past orders
Adding more international ways of payment like Visa card, PayPal, etc