From soil to deepfakes, how UAE hackathons redirect AI engineers toward solving operational problems

The experience of Bashar Nayek, a computer engineering student at the American University of Sharjah, reveals a methodological shift in how the rising generation of developers approaches AI tools. In contrast to the common pattern of starting with fascination over model sizes or massive datasets, his applied projects begin by identifying operational friction in the field first, then evaluating whether machine learning algorithms are the best tool to address it, or if the problem simply calls for direct engineering solutions.

The starting point was not generative models, but agricultural soil chemistry and field resource management.In 2022, Nayek developed an AI system that predicts mineral concentrations in soil based on chemical composition variables, temperature, pH levels, and rainfall rates to help farmers determine suitable crops, winning first place in the GITEX YouthX competition. That was followed in 2023 by a model estimating vehicle fuel consumption to improve energy efficiency, which took first place at the Dubai Roads and Transport Authority Innovation and Sustainability Hackathon, and later a browser extension developed with his team that detects deepfakes to combat digital fraud, earning the Most Innovative Project award at the InnovateX competition organised by the Telecommunications and Digital Government Regulatory Authority.

This trajectory extends beyond local tech competitions. His work was featured at the GESS Dubai classroom technology exhibition, earned the Social Endowment Award from the Indian Association Sharjah, and saw his student team represent the Middle East at the global Otis Made to Move Communities programme in 2024, securing second place worldwide with their EcoWheels sustainable transit concept. Alongside this, Nayek helped organise student platforms such as TEDx, maintaining that the practical value of innovation lies in creating opportunities and tackling real-world challenges rather than waiting for off-the-shelf solutions.

This path carries a clear lesson for the tech and education ecosystem across the Gulf, Egypt, and the Levant: the economic value of artificial intelligence does not stem from consuming generic APIs, but from directing models at specific sectoral bottlenecks such as transport efficiency, agricultural yields, and digital security. By tying student competitions to concrete operational challenges, regulatory and government bodies such as transport and telecoms authorities narrow the gap between academic research and production, pushing universities and regional startups to restructure training around problem framing and local data collection rather than theoretical training on imported algorithms.