Opening
This is what I found — what's working, where the gaps are,
and what they say about the work ahead.
I've led digital transformation across state governments, global fintechs, and diplomatic infrastructure — in LatAm, Europe, and across eight markets.
What's working
Most government digital products ship the org chart. Navigation mirrors the ministry structure, not the way citizens think about their lives.
alpha.gov.bb didn't do that. "Family, Birth and Relationships" instead of "Ministry of Home Affairs." That's a political decision before it's a design decision. Getting it right first was the right call.
Two other things worth naming:
The alpha label is a public commitment to learning, instead of a disclaimer buried in the footer.
The Tell us flow is a structured research instrument, instead of a generic feedback form.
Design direction
The typeface is Figtree — humanist, variable weight, approachable rather than institutional. The palette is national: gold (#ffc726) in the header, navy in the footer, teal (#0e5f64) on every interactive element. Black on white throughout. No visual noise.
Behind the surface: a real design token system. CSS custom properties for color, type scale, and spacing — built for more than its current scope.
The gap: legacy pages from gov.bb sit entirely outside it. The bones are right. Applying them consistently across every service page hasn't landed yet.
The core gap
The portal helps citizens find services.
It doesn't yet consistently help them complete those services.
Three findings that show where the gap is.
Finding 1
Type "passport" into the homepage search.
"We could not find any results for 'passport'."
There is a passport page on the site.
The scope limitation is disclosed after the failure, instead of before it.
Search is a trust feature. Root cause: there is no unified service index behind the portal yet. That's an architecture decision, not a content fix.
Finding 2
"Register a Birth" is labeled a digital service. The page says: "You cannot register a birth online yet."
When the label and the reality diverge, citizens stop trusting the labels — and then stop trusting the portal.
Root cause: no enforced service classification standard exists.
Finding 3
Passport. Driver's licence. Business registration. The services citizens need most are still information pages — fees, documents, a redirect elsewhere.
The navigation is user-first. The backend still reflects the org chart: TAMIS, EZPay, CAIPO remain separate ministerial islands.
The portal shows the door. Citizens still can't reliably walk through it.
What this means for the CDIO
| Finding | Root cause | CDIO responsibility |
|---|---|---|
| Search failure | No unified service index | Platform architecture |
| Label inconsistency | No classification standard | Governance |
| Thin journey wrappers | Disconnected backend systems | Interoperability |
But the CDIO role isn't only infrastructure and governance. It's also leading digital product.
The quality gap across service pages — some mature, some essentially stubs — reflects the absence of a product standard: what does a finished service page look like? Who owns it? What's the review cycle?
That's a product management question. The CDIO answers it.
How I'd sequence it
Fix trust at the front door first.
Search scope. Service labels. Service status. Fast decisions. High trust return. Unlocks everything else.
Make the flagship journeys real.
Passport. Business registration. Driver's licence. Three fully redesigned end-to-end journeys define what "done" looks like for every MDA that follows.
Instrument before scaling.
Failed searches. Handoff exit rates. Journey drop-off. A portal that can't see where it's failing can't fix itself fast enough.
How I operate
Organizational design outlasts any product.
At Guanajuato — an AI strategy engagement with Mexico's second-largest state government, 6M+ citizens — the 48% call-center reduction was the proof of concept. The reusable workflow template — adopted by three other agencies without our involvement — was the actual infrastructure for scale.
Internal capability is the defence against bad procurement.
At PayJoy — a global fintech where I ran Data & AI products across eight markets — we built AI infrastructure with internal engineering teams before contracting externally. You have to understand what good looks like before you can stop being oversold.
Redesign the process before you automate it.
At Germany's Consulate network — a foundational digital infrastructure project across ten consulates in Mexico — five days became thirty seconds, not from faster technology, but from resequencing the workflow first. Digitizing a broken process produces a broken digital process.
Two things that change what's possible
AI is a force multiplier, not a feature.
An AI-augmented team of ten can produce what previously required fifty. Autonomous agents run UI audits, review code, and monitor accessibility gaps — continuously, without adding headcount. A lean transformation team that doesn't deploy AI as a multiplier is working twice as hard for half the output.
Open-source composable infrastructure is the procurement defence.
Every monolithic vendor platform comes with hidden lock-in. Building on open, composable components means the infrastructure survives vendor failures, adapts without renegotiating contracts, and keeps taxpayer leverage intact.
Both convictions shaped how I worked across private sector and public entities. Both belong in the foundation the CDIO builds at GovTech Barbados.
Close
Architecture standards. Service delivery models. Data infrastructure. MDA capability.
These determine whether a digital transformation scales — or stays a well-designed website.
I've applied for the CDIO role. If this is worth a conversation, I'm easy to reach.