What makes an event experience launch deadline achievable?

Use AI-assisted production to extend a small team's capacity, while keeping skilled people responsible for the result. Lead developer Sergio Agosti describes a major global event build where AI coding tools acted like an extra junior developer. He estimates that AI-assisted coding was probably about ten times faster than coding by hand, and says the team would not have met the deadline without it. His account also stresses that the tools still need expert human oversight.

For G42's Intelligence Grid, a bespoke content management system supported an AI-driven video pipeline: generating images, converting them into 3D models and animating them on green screens. Production that would normally take hours per video took minutes. The team created thousands of videos and curated around 170 for the final experience, which launched across six screens at GITEX GLOBAL in Dubai. The practical approach is to combine faster generation with deliberate selection, rather than treat generated output as finished work.

Decide whether the event experience is short-lived or intended to remain in use, and judge coding compromises against that lifespan. In Sergio Agosti's account, AI-assisted development completed tasks in minutes that might otherwise have taken hours, but the initial codebase was bloated, repetitive and difficult to maintain. Generated code duplicated functionality rather than structuring it for reuse, creating technical debt: code that works immediately but slows later work.

Agosti distinguishes between compromises that may be tolerable for a short-lived project and problems that matter for a permanent platform. For a launch team, the useful decision is therefore not simply whether the experience works on opening day. It is whether the code is suitable for the experience's intended life. Expert oversight remains necessary even when faster production makes the deadline achievable.