← Back to Home

13-year-old Hyderabad student builds AI that ‘thinks before it answers’

By | Career | 27-Mar-2026 13:32:58


News Story

In a space dominated by powerful algorithms and rapid-fire responses, a 13-year-old student from Hyderabad is attempting to slow artificial intelligence down—so it can think more like humans.

Raja Dharma Tej Maddala, a Grade 7 student at Oakridge International School, has developed Raja MagRex AI™, an experimental framework designed to analyze problems from multiple perspectives before generating an answer.

“I wanted AI to think more like humans,” Raja says, outlining a vision that challenges how most modern systems operate. Unlike conventional AI models that rely on a single large engine to produce responses, his system breaks problems into layers of reasoning, simulating collaborative human thought.

At the core of Raja MagRex AI lies what he calls Artificial Civilization Intelligence—an idea inspired by how societies make decisions. Instead of a single viewpoint, the system draws on multiple “voices”, much like a council deliberating before arriving at a conclusion.

“Major decisions are rarely made by one perspective,” he explains. “They involve science, ethics, economics, and environmental thinking. I wanted to see if AI could reflect that structured collaboration.”

The architecture is ambitious. The system integrates 22 cognitive systems, 87 modules, over 100 features, and 108 personas, each designed to interpret a problem differently—through logic, science, ethics, creativity, or environmental awareness. These independent analyses are then synthesised into a unified response through an orchestration layer.

Rather than delivering instant outputs, the system is built to organize reasoning, weigh perspectives, and consolidate insights—mirroring how humans approach complex decisions.

But building such a framework at 13 has come with challenges.

“One of the biggest difficulties was understanding and organizing multiple complex ideas at the same time,” Raja says. The project required him to dive into advanced areas like AI architecture and distributed systems—largely through self-learning.

“I spent a lot of time reading research papers, technical articles, and documentation,” he adds. Experimenting with smaller prototypes helped him gradually assemble the larger system. Much of this work, he notes, happens outside school hours, during evenings and weekends.

Despite being in its early stages, Raja sees real-world potential for his AI model. He believes systems like MagRex could prove valuable in fields that demand multi-dimensional analysis—such as climate science, healthcare research, and large-scale engineering challenges.

“These areas involve scientific data, ethical considerations, and practical constraints all at once,” he says. “A structured AI system could help decision-makers arrive at more balanced and informed solutions.”

For now, his focus remains on completing the first working version of the system and testing how its modules interact. Beyond that, he hopes to evolve it into a research platform exploring new AI architectures.

“My long-term goal is to build AI that helps humanity solve complex global problems responsibly,” he says.

At an age when most students are still discovering their interests, Raja is already redefining how intelligence—human or artificial—can function. His work is not just about building smarter machines, but about making them more thoughtful.

“I wanted AI to pause, consider, and then respond,” he says. “That’s the kind of intelligence I want to create.”