Imagine leading a company with customers all over the world. You walk into your office on a Monday morning, coffee in hand. You open your laptop and type: “Show me the number of my business customers from Zimbabwe.” In seconds, the number appears: 647!
Well, that’s the vision behind Query_me, a brand-new tool created by Realeboha Letsae and Relebohile Letsae, two third-year students (and twins) at the National University of Lesotho (NUL), studying Computer Science and Statistics.
______________
Brought to you by NUL Innovation Hub: Where Academia Meets Industry!
No calling the IT department.
No waiting for the data analyst to get back from lunch.
No trying to remember that long, complicated SQL command from last year.
Just a straight answer. From your own system. In plain English.
And yes, if you’re wondering…the developers are indeed twins.
“We just wanted to make life easier for non-techies,” they said as they introduced their work.
“Most people in business know what they want to ask,” says Relebohile, “but they don’t know how to speak the language of digital databases.”
Her twin, Realeboha, adds:
“So they end up sending requests to IT guys, who are often overwhelmed. We thought; what if business leaders could just type their questions in English and get an answer straight from the company system?”
And that’s exactly what they’ve built.
What Is Query_me?
Query_me is like a translator between people and data. You type your question in plain English, and the system instantly converts it into SQL; the language computers use to get answers from stored information. It runs the command, fetches the data, and displays the result.
Just like that.
Let’s say a CEO wants to know:
“Which products sold the most in May?”Query_me translates that into a valid SQL query, runs it, and gives the answer. All in seconds.
In one case, they tested it using a sample music album database (called the Chinook database), but the twins say the tool can be adapted to any business data—sales, inventory, HR, customer records…you name it.
“The point is speed. And independence,” they said.
The problem is, they say: waiting for data slows down business.
In many companies, managers and executives without a technical background often wait hours, or even days, for developers to create simple reports.
“That’s a waste of time,” says Relebohile. “A manager might wait days just to find out how many customers they had last month.”
With this tool? No coding. No confusion. Just answers.
So, How Did They Build It?
Let’s get a little techy (just for a moment).
The system is made up of these parts:
• Frontend (React): The user-friendly interface where users type questions and view results.
• Backend (Flask): The engine that links the interface, AI, and data behind the scenes.
• Database (MySQL): Where all the information lives.
• de-fog-sqlcoder-7b: The smart brain from Hugging Face that converts English into SQL.
• Cloud Runtime (Google Colab): Where the system runs, so it doesn’t depend on your local computer.
And yes, all of this was built and deployed by two university students…while still attending classes and doing assignments!
Like most great ideas, this one didn’t come from a lecture hall.
“It started as a small experiment,” they said. “We just wanted to see if we could get a model to understand natural language and turn it into SQL. Then we realized; this could actually help real people.”
So they turned the experiment into a fully working system. They tested it, hosted it, and made it live.
You can even check it out here:
🔗 https://github.com/Relebohile-L/query_me
So, What’s Next?
The system is working, it is working well. Now the twins want to share it with the world.
They hope companies, government offices, and even small businesses will find it useful, especially those with limited tech teams.Even small businesses deal with data.
That’s the heart of Query_me; making data talk back. Quickly. Clearly. And in plain English.
___________
Brought to you by NUL Innovation Hub: Where Academia Meets Industry!
