CDB PROPEL Is a $5 Million Bet on Caribbean Small Business. AI Is the Multiplier It Needs

On July 8, 2026, two press releases went out roughly a thousand kilometres apart that were, in substance, about the same problem. In Castries, Saint Lucia, CARICOM Heads of Government wrapped their 51st Regular Meeting of the Conference with a joint statement on the cost of living, the kind of statement regional summits produce every year, except this one arrived with unusually specific commitments: cutting freight and logistics costs, diversifying import sources, testing a Trinidad and Tobago ferry as a cheaper route for inter-island trade, and expanding consumer technology that lets a shopper compare supermarket prices before they check out. In Bridgetown, Barbados, the Caribbean Development Bank announced CDB PROPEL, a new chapter in the Bank's technical assistance mandate for micro, small, and medium-sized enterprises, retiring the Caribbean Technological Consultancy Services Network after more than four decades of service to the Region's entrepreneurs.

Neither announcement made international headlines. Neither one should be read in isolation either. A region that is worried enough about the cost of living to convene a private sector council on freight costs is a region where the small businesses selling to those same households are also being squeezed, on the supply side, by the identical forces: import costs, thin margins, and limited access to the kind of planning tools that larger competitors already use. CDB PROPEL is the institutional response on the business side. What it becomes over the next three years depends heavily on whether AI gets built into its design now or gets treated as an optional add-on later, the way too many well-funded technical assistance programmes have treated it before.

What CDB PROPEL Actually Replaces

The Caribbean Technological Consultancy Services Network, known across the region simply as CTCS, has been the workhorse of Caribbean MSME support since the early 1980s. It connected entrepreneurs with consultants, funded diagnostic studies, and helped thousands of small operations formalize bookkeeping, adopt new equipment, and reach markets outside their home island. It was not glamorous work. It was also, by any reasonable measure, effective at the scale it operated. Between 2015 and 2025 alone, CDB approved approximately USD 11.8 million in technical assistance that touched more than 10,000 MSMEs and entrepreneurs, spanning business management, market access, renewable energy adoption, climate resilience, youth entrepreneurship, financial inclusion, and women's economic empowerment.

CDB PROPEL is not a rebrand of that legacy. It is a deliberate acknowledgment that the entrepreneur CTCS was built to serve in 1983 is not the entrepreneur CDB PROPEL exists to serve in 2026. The Bank's own language on the launch put it plainly: the entrepreneur today is more connected, more ambitious, and competing on a global stage, and the programme's renewed commitment is to make sure that entrepreneur has the expertise, networks, and opportunities needed to compete, adapt, and thrive. CDB has allocated USD 5 million under the 11th cycle of its Special Development Fund for MSME technical assistance running from 2025 through 2028. That is a real, if modest, sum for a mandate covering every CDB member country, from Jamaica and Guyana down to the smallest OECS states.

The scale of what CDB PROPEL is responsible for is worth stating plainly, because it explains why the design choices made in its first year matter far beyond the programme's own budget line. MSMEs account for 70 to 85 percent of businesses across the Caribbean, contribute between 60 and 70 percent of regional GDP, and provide roughly half of regional employment. A technical assistance programme touching this segment is not a niche development intervention. It is close to the single most consequential lever CDB has for regional economic resilience, which is precisely why the tools it teaches in 2026 need to be the tools that matter in 2026, not a modernized version of what CTCS taught a decade ago.

The Cost-of-Living Crisis Is a Small Business Story Too

The framing coming out of Saint Lucia treated the cost-of-living crisis mostly as a household affordability problem, and it is that. But every dollar a household loses to a higher grocery bill traces back through a supply chain that runs through a wholesaler, a distributor, and very often a small retailer who is absorbing the same freight, fuel, and import cost increases that CARICOM's private sector council spent four days discussing. The independent hardware store in Portmore, the family-run pharmacy in Castries, the craft exporter shipping out of Georgetown: each of these businesses is squeezed by the same import cost pressure hitting households, with none of the tools a larger regional retailer has to model that pressure, forecast it, or price around it in advance.

This is the part of the cost-of-living conversation that gets less attention than it should. A household cutting back on groceries is a visible, immediate political problem. A small business quietly losing margin on every order because it cannot forecast demand accurately, cannot time its inventory purchases against currency and freight movements, and cannot access financing without collateral it does not have, is a slower, less visible version of the same crisis. It shows up eighteen months later as a shuttered storefront, not as a headline. CDB PROPEL is, whether the language says so explicitly or not, a programme aimed squarely at that second, quieter half of the cost-of-living story.

The private sector council's own proposals point toward the same conclusion from a different angle. Reducing freight costs, diversifying import sources, and piloting a cheaper inter-island ferry route are all supply-chain interventions. So is helping ten thousand MSMEs forecast demand accurately enough that they stop over-ordering perishable stock or under-ordering fast-moving goods. The region is, in effect, running two parallel supply-chain resilience programmes this month, one aimed at the shipping lanes between islands and one aimed at the businesses receiving what those ships carry. Neither works at full strength without the other.

Where AI Belongs Inside CDB PROPEL, Concretely

It would be easy, and wrong, to write this as a generic call for CDB PROPEL to "embrace AI" without specifying what that means for a business support organization designing a curriculum this year. The useful version of this argument is concrete. There are four categories of AI tool that fit the budget, skill level, and daily reality of the businesses CDB PROPEL exists to serve, and each maps directly onto a gap the programme's own stated goals already identify.

The first is cash flow and demand forecasting. Most MSME owners in the Caribbean are running inventory and cash decisions on instinct and a notebook, not because they lack the discipline to do better but because the tools that do better have historically required a level of technical skill and cost that put them out of reach. AI-powered forecasting tools have collapsed both barriers in the past two years. A shop owner can now feed a few months of sales data into a low-cost tool and get a demand forecast that flags a stock-out or a cash crunch two weeks before it happens instead of learning about it when the shelf is already empty. This is precisely the kind of business management support CTCS built its reputation on. AI does not replace that mission. It makes the same mission achievable for far more businesses per consultant-hour.

The second is bookkeeping and financial formalization. A meaningful share of Caribbean MSMEs operate with financial records that exist mostly in a shoebox of receipts and a proprietor's memory, which is exactly the condition that keeps them from qualifying for formal financing on reasonable terms. AI-assisted bookkeeping tools, the kind that read a photographed receipt and categorize it automatically, have made basic financial formalization achievable in weeks rather than the months a manual consulting engagement used to require. This is not a hypothetical use case. It is one of the most immediately deployable AI applications for exactly the population CDB PROPEL is targeting.

The third is export and compliance documentation. A craft exporter or agro-processor trying to reach a new CARICOM market, or a market beyond the region, faces a wall of paperwork, standards documentation, and compliance requirements that has historically required either an expensive consultant or months of a proprietor's own time to navigate. AI tools that draft, check, and structure this documentation against known regulatory templates can cut that timeline meaningfully, and this is a use case that connects directly to the private sector council's own push to diversify trade and reduce the friction of moving goods around the region.

The fourth is credit-risk scoring for alternative lending. This is the least visible but potentially highest-impact application. Formal lenders across the Caribbean still rely heavily on collateral and credit-history requirements that exclude a large share of small, informal, and women-led businesses, the same population CDB's technical assistance data shows has been a consistent priority for the Bank over the past decade. AI-driven credit scoring, which builds a risk profile from transaction history, mobile payment patterns, and other alternative data rather than collateral alone, has already proven itself in financial inclusion programmes across Africa and South Asia. There is no structural reason it could not do the same work for a Caribbean market that shares many of the same informal-economy characteristics.

PricePal and the Consumer Side of the Same Idea

The CARICOM summit's own reference to consumer price comparison technology is a useful, concrete example of the same principle applied to households instead of businesses. PricePal, a grocery price comparison app that began expanding its national rollout in the Bahamas in March 2026, lets a shopper compare prices across participating retailers before they buy, turning what used to require walking between three supermarkets into a five-second check on a phone. It is a small tool. It is also exactly the category of application CARICOM leaders pointed to when they discussed technology solutions for the cost-of-living crisis, and it demonstrates something CDB PROPEL's designers should take seriously: the AI and data tools that actually change a Caribbean household's grocery bill tend to be modest, purpose-built, and cheap to deploy, not the kind of large-scale enterprise system that requires a multi-year procurement process.

That is the right model for CDB PROPEL to borrow. The programme does not need to fund the development of a single, region-wide AI platform for MSMEs, an approach that would burn through a significant share of its USD 5 million budget before touching a single entrepreneur. It needs to build AI literacy and access to existing, affordable tools into its technical assistance curriculum from day one, the same way CTCS spent 40 years building basic business management literacy into that curriculum. The tools already exist. What is missing is the systematic effort to put them in front of ten thousand business owners who do not yet know they are affordable, or that they exist at all.

The Risk of Getting This Wrong

Massy Holdings and ANSA McAL are running AI pilots across their retail, insurance, and logistics operations in Trinidad. Major regional banks are deploying AI for fraud detection and underwriting. Larger tourism operators are testing AI-driven pricing and demand forecasting. None of this is controversial or new to readers of this blog. What matters for CDB PROPEL is the gap this creates if it is left unaddressed: a two-tier Caribbean private sector where large, well-capitalized businesses compound an AI-driven cost and forecasting advantage year over year, while the MSMEs that employ half the region's workforce keep running on instinct and a notebook because nobody built AI literacy into the technical assistance they actually receive.

That gap does not stay contained to competitiveness statistics. It shows up as the independent shop that cannot match a larger retailer's pricing because it over-orders perishable stock every month. It shows up as the small exporter who loses a contract because a competitor's AI-drafted compliance paperwork moved through customs three weeks faster. It shows up, eventually, as exactly the kind of shuttered storefront and lost employment that both CDB PROPEL and the CARICOM cost-of-living package are trying to prevent. A technical assistance programme that modernizes everything about how it delivers support except the tools it teaches has done half the job CDB PROPEL was designed to do.

What This Requires, and Who Has to Move

None of this requires CDB to become a technology company. It requires three specific decisions inside CDB PROPEL's first year of operation. First, build a baseline AI literacy module into every technical assistance engagement funded under the programme, not as an elective add-on but as a standard component, the same way basic bookkeeping became a standard component of CTCS engagements decades ago. Second, partner with regional AI providers, including Caribbean-founded companies that already understand the specific constraints of a small, informal-economy business, rather than defaulting to global enterprise vendors whose tools are priced and designed for a different market. Third, measure adoption. CDB's own reporting on CTCS tracked reach in the tens of thousands over a decade. CDB PROPEL should set an explicit target for how many of those MSMEs are using at least one AI tool in their operations by the end of the SDF 11 cycle in 2028, and report against it the way it reports against every other metric in its technical assistance portfolio.

Business support organizations across the region, the chambers of commerce, the small business development agencies, the entrepreneurship hubs that CDB PROPEL is designed to strengthen, do not need to wait for a formal CDB mandate to start this work. A chamber of commerce in any CARICOM member state can run a half-day AI literacy session for its members this quarter using tools that already exist and cost less than a lunch meeting per participant. The MSME owners themselves do not need to wait either. A twenty-dollar monthly subscription to a forecasting or bookkeeping tool is a smaller line item than most businesses already spend on far less useful software, and the payback period, measured in avoided stock-outs and recovered margin, is usually a matter of weeks.

"CTCS spent 40 years teaching Caribbean entrepreneurs to keep a proper ledger. CDB PROPEL's job is teaching them to forecast with a tool that a ledger alone can never give them. The mission has not changed. The toolkit finally has, if we choose to use it." - Adrian Dunkley, AI Boss

I founded StarApple AI on the belief that the Caribbean's small businesses are not a footnote to the region's AI story, they are the story. A sovereign AI model or a headline-grabbing enterprise deployment matters, but it is the hardware store, the agro-processor, and the craft exporter, multiplied across every CARICOM member state, that actually determine whether the region's cost-of-living crisis gets better or worse over the next three years. CDB PROPEL has the funding, the institutional relationships, and forty years of earned trust with exactly this population. What it does with the next three years, and whether AI is part of the curriculum from the start or an afterthought bolted on in year two, will decide whether this becomes the programme that closed the gap between Caribbean MSMEs and their larger competitors, or the programme that modernized everything except the one tool that would have mattered most.