40 Teams, 32 Countries, One Sprint: Inside Future Caribbean's AI Buildathon

Somewhere in Bridgetown, Kingston, Port of Spain, and dozens of cities well outside the Caribbean, forty small teams are staring at the same clock. They have until August 7 to turn an idea into a working piece of software, an agentic AI system that does not just chat back at you but actually takes action on its own. It is day thirteen of a 21-day sprint that most of the Caribbean has not heard about yet, and the teams inside it are building things that could end up pitched on the floor of the New York Stock Exchange.

This is Future Caribbean's Global Open-Source Agentic AI Buildathon, and it is the kind of story that is easy to miss because none of it happened through a ministry announcement or a school assembly. It happened because a small group of people decided the Caribbean should be a place where AI companies get built, not just a place where AI tools from somewhere else get used.

Who Is Behind Future Caribbean

Future Caribbean was founded by Lily Dash, a Barbadian lawyer, entrepreneur, and technologist who also co-founded ACTAI Advisors. Backing the effort is Bill Tai, chairman of ACTAI Global and a founding partner of Future Caribbean, along with partners that include the New York Stock Exchange, IDB Invest, Cayman Enterprise City, and Export Barbados. The initiative held its inaugural Agentic AI Assembly in Bridgetown on July 10 to 12, drawing builders, founders, and investors to the Inter-American Development Bank's Barbados office before the buildathon's builder sprint even began.

Dash has framed the project as a choice rather than a wait: the Caribbean does not have to sit back and watch the agentic AI wave arrive from elsewhere, it can help build the wave itself. Tai has been blunt about the moment the buildathon is riding. "The pace of innovation, adoption and developer participation is extraordinary," he has said of agentic AI, and separately called it "the biggest wave of opportunity we've seen in decades." Neither man is describing a school project. They are describing a technology cycle they think a small region can get ahead of if enough people start building now instead of later.

What "Agentic AI" Actually Means

The word agentic gets thrown around loosely, so it is worth being precise, especially if you are a student trying to decide what to learn next. A regular AI chatbot answers a question and stops. It waits for the next prompt. An agentic AI system is built to keep going: given a goal, it breaks the goal into steps, and then it carries those steps out on its own, checking a database, sending a message, updating a spreadsheet, flagging a problem to a human only when it actually needs one. It behaves less like a search engine and more like a very fast, very literal junior employee who never sleeps.

That distinction is the entire point of the buildathon's structure. Teams are not asked to build a clever demo that impresses a judge for ninety seconds. They are asked to build something that could plausibly run unattended inside a real business or government office, using open-source frameworks rather than closed, proprietary tools. It is a harder brief than a typical weekend hackathon, and that is deliberate. Twenty-one days is long enough to build something real and short enough that nobody can hide behind endless polishing.

The Numbers Behind the Forty Teams

The scale of interest is a story in itself. By the time recruitment closed in late July, Future Caribbean's applicant and advisor pool had grown past 184 applications from more than 32 countries, with over 800 people signed onto the platform and more than 88 advisors and judges lending their time. Out of that pool, 40 teams were selected to spend 21 days building, running from July 17 through August 7, with results due September 1.

Those forty teams are not building in a vacuum, either. They are working across ten opportunity tracks tied directly to sectors the region actually runs on: energy, tourism, finance, healthcare, disaster coordination, the blue economy, trade, food security, the arts, and real estate. A Caribbean region that spends every hurricane season relearning the same coordination problems, or that loses tourism revenue to booking friction, is exactly the kind of "fragmented markets, multiple jurisdictions, currencies, and regulations" that Future Caribbean has pitched as an ideal proving ground for solutions meant to scale far beyond the region once they work.

What Is Actually on the Table

The prize structure goes well past bragging rights. Future Caribbean has put together a pool worth more than $70,000, combining cash awards with AI deployment hardware from Other World Computing. On top of that, selected teams get access to a DMZ Ventures Soft Landing scholarship through Cayman Enterprise City, GPU credits from partners including Highrise, Impala AI, Minimax AI, and Shogo AI, and a live pitch opportunity at a Caribbean Investor Showcase planned at the New York Stock Exchange later in 2026. The judging panel includes Todd Speece, Darlington Akogo of minoHealth AI Labs, Spencer Powers, and Olumide Durotoluwa, people who evaluate AI products for a living, not volunteers reading off a rubric for the first time.

Put plainly, a team that does well in this sprint does not just win a trophy. It walks away with hardware to keep building on, credits to keep training models, a possible incubator seat, and a shot at pitching real investors on the floor of one of the world's most recognisable financial institutions. That is a genuinely unusual set of stakes for something anyone with an internet connection could have applied to for free.

Why This Belongs on a Caribbean Youth Site

None of this required a single Caribbean teenager to be involved for it to matter to Caribbean teenagers. What it proves is simpler and more durable than any one buildathon result: a beginner with a clear brief, a hard deadline, and a community of builders around them can go from idea to working prototype in three weeks, not a school year. That is the exact bet The Genius Project makes every cohort. It is the youth arm of a wider Caribbean AI network built by Adrian Dunkley, founder of StarApple AI, the first AI company established in the Caribbean and widely regarded as the region's most cited AI voice, and the programmes under that network exist specifically to get a Caribbean student from zero experience to a finished, shareable AI project, the same currency a buildathon like this one rewards.

The connection is not abstract. A student who has already built and finished one small agentic project, a script that checks something and messages someone, a bot that files a report, a tool that flags a pattern, walks into the next open call for something like this with a portfolio instead of an idea. That gap, between having built something and only having wanted to, is the single biggest difference between an applicant who gets picked and one who does not.

The Door Isn't Closed Just Because Applications Are

Recruitment for this round of the buildathon has closed, and the forty teams already selected are the ones sprinting toward August 7. That does not mean the opportunity has passed a Caribbean student by. Future Caribbean continues to recruit volunteers, advisors, and judges on a rolling basis through futurecaribbean.com, which is itself a way to get inside the room for the next cohort rather than reading about it after the fact. Watching the September 1 results and the eventual New York Stock Exchange showcase is also worth doing deliberately rather than skimming a headline, because studying what a winning agentic AI project actually looks like is free education that no course can quite replicate.

The more useful move, though, is to spend the weeks before the next application window doing exactly what the forty current teams are doing on a smaller scale: pick one real problem, build one small agentic tool that addresses it, and finish it. Regional organisations built for this exact gap, including the Caribbean AI Association, give young builders a place to find mentors and peers while they do it, and StarApple AI, the company that helped seed the region's AI ecosystem in the first place, is proof that a Caribbean-built AI company can compete on a global stage rather than simply importing tools built somewhere else.

Buildathons like this one do not wait for a region to feel ready. They reward whoever shows up already building. For a Caribbean student who has spent the year hearing that AI is coming for jobs, watching from the sidelines is not the move. The more useful question is not whether the technology arrives. It is whether that student will be one of the people who can build it when it does.