The Dominican Republic is the largest economy in the Caribbean, at roughly US$120 billion in GDP. It receives more than 10 million visitors a year, runs one of the region's biggest free zone manufacturing sectors, exports cocoa and coffee, and takes in about US$10 billion a year in remittances. For its size, it spends little on AI.
That gap is the opportunity. Each of the five sectors below already produces data every day, and each has an organisation that could run a pilot within a year. I have left out the national-strategy questions and kept to where a company or cooperative can start.
Where the Dominican economy earns its money
| Sector | Scale | Data it already holds |
|---|---|---|
| Tourism | About 17% of GDP including indirect effects; 10 million+ visitors a year | Bookings, rates, occupancy, guest reviews |
| Free zone manufacturing | About US$7 billion in exports; 170,000+ workers | Quality inspection records, machine sensor logs |
| Export agriculture | Cocoa exports about US$200 million a year | Cooperative purchase records, export and certification paperwork |
| Banking and credit | Roughly half of adults without a formal account | Remittance receipts, utility and mobile payments |
| Remittances | About US$10 billion a year, the largest foreign currency inflow | Transaction logs at transfer operators and banks |
1. Resorts: demand forecasting and pricing
Large all-inclusive resorts in Punta Cana already run occupancy and pricing models. The gap is in how many signals feed them. Forecasting models that also use flight searches, source-market economic data, weather and competitor rates can set room prices and plan staffing, food and supply orders more accurately.
How to size it: multiply your annual room revenue by the revenue-per-available-room gain you expect. A 3% gain on US$50 million of room revenue is US$1.5 million a year. Ask any vendor for the RevPAR change at comparable properties, measured over at least a full year.
Where to start: the property and revenue management systems that resort groups already license often include forecasting modules that nobody has switched on. Check your current contract before buying anything new.
2. Free zones: visual inspection and predictive maintenance
Medical devices are the largest and most technical part of the free zone sector, and the plants already meet US FDA quality rules. AI does not replace their quality systems. Cameras trained on production-line images can inspect every unit at line speed, and models reading machine sensor data can warn of failures before they stop a line.
How to size it: take last year's scrap and rework cost plus the hours of unplanned downtime on one line. That total is the ceiling for a pilot's return on that line.
Where to start: the plant's own quality engineers, with a vision-inspection vendor, on a single line for 90 to 120 days. For a sector-wide pilot, the Asociación Dominicana de Zonas Francas (ADOZONA) and the Consejo Nacional de Zonas Francas de Exportación are the natural conveners.
3. Cocoa and coffee: crop monitoring and traceability
The Dominican Republic is among the world's largest exporters of organic cocoa, and the price premium depends on consistent quality and proof of origin. Satellite crop monitoring can flag drought stress and disease risk before symptoms show. Digital traceability records each lot from farm to container, which is what buyers paying organic and ethical premiums ask for.
How to size it: on about US$200 million in cocoa exports, a 15% yield gain plus a 10% quality premium would add roughly US$50 million a year. Treat both percentages as targets for a pilot to test.
Where to start: CONACADO, the national confederation of cocoa growers, has the cooperative structure and buyer relationships. A pilot with 500 farmers over two harvests would show whether the premium holds.
4. Banking: credit scores for people without a credit file
Roughly half of Dominican adults have no formal bank account, yet many receive remittances, pay utility bills and run small businesses. Alternative-data scoring uses those payment histories to assess credit risk for people with no bureau file. Kenya's M-Shwari and Mexico's Konfío are working examples.
How to size it: reaching one million adults with an average facility of US$2,000 would put US$2 billion of new credit into the economy. Default rates in the pilot decide whether that is sustainable.
Where to start: Banco Popular, Banco BHD and the larger microfinance institutions have the reach. A supervised pilot of 10,000 applicants, agreed in advance with the Superintendencia de Bancos, is achievable within six months.
5. Remittances: fraud and diversion checks
Most of the US$10 billion in annual remittances passes through a small number of transfer operators and banks. Anomaly detection on those transactions can flag account takeovers, mule accounts and unusual routing patterns, which also helps with anti-money-laundering reporting.
How to size it: start from the operator's current fraud losses and chargebacks, and the staff hours spent on manual transaction reviews. Those two numbers set the budget ceiling for a detection pilot.
Where to start: the compliance team of one transfer operator, working from a year of historical transactions with confirmed fraud cases labelled.
How to read these numbers
The sizing figures above are worked examples, not forecasts. Replace my percentages with the baseline from your own property, line or portfolio, and ask every vendor for measured results from a comparable site before you believe a gain.
What a Dominican company should do in the next 90 days
- Chief executive, week 1: pick one problem you already measure, such as RevPAR at one property, scrap on one line or default rates in one branch. Write down this year's figure as the baseline.
- Operations or IT lead, weeks 1 to 4: find where the data lives (PMS, quality records, export paperwork) and check you hold at least 12 months of it in usable form. If you do not, the first project is cleaning the data.
- Legal lead, before any data leaves the building: check the plan against Ley 172-13 on personal data and, for banks, the Superintendencia de Bancos' rules on outsourcing and data.
- Project owner, weeks 5 to 12: run a 60-day pilot on one property, one line or one cooperative chapter. Use a cloud platform with Spanish-language support instead of building from scratch.
- Chief executive, day 90: compare the result with the baseline. If it moved, fund the next site. If it did not, write down why (usually data quality or the wrong question) before trying again.
Frequently Asked Questions
Which Dominican law governs personal data used in AI projects?
Ley 172-13, the Dominican Republic's law on the protection of personal data, passed in 2013. It sets rules on consent and on how credit information bureaus handle personal data, which matters for any alternative credit-scoring project. Banks should also follow the Superintendencia de Bancos' rules on technology risk and outsourcing. Get local legal advice before sending customer data to a foreign cloud provider.
Is there public funding or support for AI pilots in the Dominican Republic?
Check with the Ministerio de Industria, Comercio y Mipymes and with export promotion agency ProDominicana, which run programmes for small firms and exporters. Multilateral lenders, including the Inter-American Development Bank, finance digitalisation projects in the country. Cooperatives can also ask their international buyers, some of whom co-fund traceability systems because they need the records.
How do I check a vendor's claim about revenue or yield gains?
Ask for the before-and-after figures from a named client property or farm, the length of the measurement period, and what else changed at the same time. A year-long comparison against a similar site that did not use the tool is the standard to aim for. Treat a percentage with no baseline, no period and no comparison as marketing.
Can small Dominican firms use AI without a data science team?
Yes, for the first step. A general assistant such as ChatGPT, Claude or Gemini on a paid plan, at around US$20 a month per user, can draft Spanish and English guest replies, summarise supplier contracts and analyse a sales spreadsheet. Forecasting, visual inspection and credit scoring need a specialist vendor or hire. Do not paste customer personal data into a consumer AI plan.
Where does the Dominican Republic's skills gap show up first?
In data engineering, the unglamorous work of getting records out of hotel, factory and bank systems into a usable form. Model-building skills are easier to buy from abroad than people who understand the local systems. Firms can partner with universities such as INTEC or PUCMM for interns, and recruit from the Dominican diaspora in the United States, which includes experienced engineers.