Nilesh Ramgolam: A Mauritian at the Forefront of International Innovation

Nilesh Ramgolam, a Mauritian researcher based in Adelaide, Australia, has recently achieved two significant milestones in his academic journey, marking a major advancement for both himself and Mauritius on the international stage.

On December 6, 2025, Nilesh Ramgolam received a national award at the GovHack 2025 ceremony, the largest government open data hackathon in the Southern Hemisphere, held in Sydney. His project, 'Community AI Agents: Bridging Service Access Gaps', aims to develop artificial intelligence agents to help citizens navigate government services more easily, and it was awarded first prize.

This system leverages his expertise in AI tailored to preferences, with a particular focus on community impact and accessibility to essential services. This victory is especially significant for Nilesh Ramgolam, as it proudly represents Mauritius in an international innovation competition.

A week earlier, Nilesh Ramgolam also presented his research at DICTA 2025, a premier conference on computer vision and machine learning. His paper, 'Learning to Defer to a Population with Limited Demonstrations', proposes a new approach to AI, aiming to determine when systems should act autonomously and when they should call on human expertise. This work has crucial implications for sensitive fields such as medical imaging.

This impressive journey began in Mauritius, where Nilesh studied at the Royal College Curepipe and John Kennedy College, before pursuing his studies at the University of Adelaide, where he received several prestigious scholarships. Today, he is completing an MPhil in artificial intelligence while teaching and actively contributing to scientific innovation.

In his future projects, Nilesh Ramgolam aims to pursue a PhD in computational drug discovery, with the ambition of revolutionizing drug design through AI. His journey inspires young Mauritians to pursue careers in science and technology, fields where they can excel on a global scale.