Every International Women's Day brings the graphics, the quotes and the pledges to "amplify women's voices". The commitment that counts happens on the other 364 days, in hiring and promotion decisions and in the meetings where a qualified woman is passed over because she did not fit someone's picture of a technologist. This piece is a thank-you to specific people, and a list of the gaps in Caribbean AI that I think employers and funders can close this year.
To the women of StarApple AI and our companies
To every woman on the teams at StarApple AI, Maestro AI Labs, the Section 9 AI Lab, IMPACT AI Lab, SportsBrain, AppleSeed Trainlytics, SuReal AI and The Genius Project: these organisations work because of you, in the daily sense of showing up and building things.
You brought rigour to our research and imagination to our products. You asked whether our systems served the communities they were built for, and you pushed back when the data told a different story from the one we expected. You stayed late, shipped, and then explained to partners and clients what we had built and why it mattered.
Caribbean tech has long been dominated by men, on conference stages, in boardrooms and at hackathons. The women in our companies built anyway, and I am grateful for it.
To the women in my life
Teachers saw something in a neurodivergent boy that the school system was not designed to see. They explained things a different way when the standard explanation did not land, and stayed after class with a patience no job description required.
The women in my family, with roots across cultures and continents, held the household together and showed me that strength and warmth sit comfortably in the same person.
Colleagues and partners backed work that was not yet proven, in a Caribbean AI sector most people could not yet see. They made introductions, sent the email and put their own credibility behind mine when it mattered. Thank you, all of you.
Why diverse teams build better AI
AI learns from data, and data reflects who was counted. Cardiac research has historically under-represented women, so models built on it can miss symptoms that present differently in women. In 2018 Reuters reported that Amazon scrapped an experimental recruiting tool after finding it downgraded CVs containing the word "women's", because it had learned from ten years of mostly male hires.
Teams carry the same blind spots into design. A team without women asks fewer questions about how a system could fail women. That is the core of what I call Artful Intelligence: diverse teams catch more failure modes, so they ship better systems.
Where Caribbean AI still loses women
- Progression. My free training sessions, now in their seventh year, draw strong female attendance, but fewer women move on to advanced research and product roles. Caregiving, money pressure, few senior women to learn from and old ideas about who belongs in technical work all play a part.
- Capital. Venture funding for women-led AI startups in the Caribbean is thin. The 14West Fund is one small answer, and it needs to be bigger and to have company.
- Policy rooms. AI task forces, advisory committees and standards bodies still have too few women, because the pipelines into those rooms have gaps, not because qualified women are missing.
I am least sure about the causes of the progression gap. My sense comes from watching my own classes, not from tracked data, and it may be that the barriers differ between islands and income levels more than I assume. Anyone running training should measure it rather than take my word.
What Caribbean employers and funders can do this year
- Heads of AI and data teams: track, by gender, who applies, who is hired and who is promoted for the last three years, and publish the numbers internally before setting any target.
- Training providers, including us: record who finishes introductory courses and who reaches advanced ones, and offer at least one advanced cohort at hours that work for parents.
- Hiring managers: put at least one woman on every technical interview panel and score candidates against written criteria agreed before interviews start.
- Investors and development banks: report what share of your AI deal flow and funding goes to women-led companies, and set aside a first-cheque allocation for them.
- Ministers appointing AI task forces: require balanced shortlists for every seat and publish the membership.
To the young women of the Caribbean: if you are in Jamaica, Trinidad, Barbados, Guyana or anywhere in this region and wondering whether AI is for someone like you, it is. The AI that serves the Caribbean best will be built by people who know its dialects, culture and economics. The technical skills can be learned; I have spent seven years proving that in free classrooms and will keep going. Come build with us.
Frequently Asked Questions
Where can Caribbean women find AI training and community?
Start locally with The Genius Project, university continuing education at UWI and UTech, and free online courses you can audit. Women in Data Science (WiDS) Worldwide, which began at Stanford, supports local events that anyone can organise, so a Caribbean chapter can be started by one committed volunteer. Pair any course with a project on local data to build a portfolio.
How can a company check whether its AI system treats women unfairly?
Compare the system's outcomes and error rates for women and men on real cases, such as approval rates in a credit model or pass rates in a CV screener. A common rough screen is the four-fifths rule used by US employment regulators: if one group's selection rate is below 80 percent of another's, investigate. Then check which input features drive the gap.
What funding is available for women-led AI startups in the Caribbean?
Options are limited but growing. The 14West Fund invests in Caribbean startups, national development banks such as the Development Bank of Jamaica run programmes for small businesses, and the World Bank's Women Entrepreneurs Finance Initiative (We-Fi) funds programmes through partner institutions. Ask each funder directly what share of its portfolio is women-led; the answer tells you how seriously it takes the question.
What keeps women in technical roles once they are hired?
Fair treatment, mostly. The Kapor Center's 2017 Tech Leavers Study found unfair treatment was the most common reason people left tech jobs, ahead of better offers elsewhere. The practical fixes are transparent promotion criteria, a named senior sponsor for each high performer, and flexible hours that do not count against someone at review time. Exit interviews that ask why people leave will show which applies in your company.