For many service businesses - think freelancers, plumbers and other tradespeople, my barber - the calendar is the most important thing for their operations. Every empty slot is lost revenue, every booked slot represents a sale. While many businesses understand the relationship between their calendar and their revenue, there is still a big reliance on gut-feel to understand the impact day-to-day, especially if they primarily accept cash where there is no clear electronic paper trail.

This prompted me to consider whether Cari and its focus on micro/small businesses in emerging markets, where so much is manual, could potentially deliver a simple, accessible way to leverage the calendar as a simplified proxy for revenue.

The Feature: Mapping Time to Money

Appointments and calendar functionality have long been on my radar for Cari, I even explicitly mentioned it back in November 2025. It’s not hard to see why, I’ve consistently received feedback that while expenses are important, anything that touches revenue is more immediately valuable. Cari already supported basic revenue logging functionality.

This works, but requires the business to remember to log each transaction as it happens. Something I see very few cash-based businesses do, especially if it’s a busy day with customers constantly arriving waiting for their turn. So I figured, “what would be the simplest way to approach this?” and this led me down the path on this calendar-to-revenue feature. Here’s what it actually does in practice:

Businesses link their Google Calendar via their existing WhatsApp conversation with Cari. With that alone, they can already ask detailed questions about their schedule.

But on top of that, because Cari already supports basic revenue functionality, they can now ask Cari to identify any events on their calendar that could represent revenue and Cari will help them log that revenue.

I want Cari to stay out of the way as much as possible. It’s there when a business needs it, but invisible when they don’t. Anything that impacts their real-time workflows with paying customers could lead to me actually costing them money. As such, a key tenet for building Cari was ensuring businesses could use the tool at their convenience retroactively, which could be at the end of the day, or week, or whenever.

Not disrupting their existing flows for managing their schedule was important as well, so it doesn’t require them to use Cari to create/update/delete events, although they can.

Restraint Is Now the Hard Part

The calendar-to-revenue feature is simple. That’s not a limitation, it’s an intentional decision. There’s no shortage of feature rich appointment systems out there already if they want one.

In my experience, restraint has often been a by-product of juggling too many roadmap items with too few technical resources. With AI having proven itself as an undeniable unlock in product development, it is easier than ever to overbuild or deliver frankenstein products. It’s no longer “can we?” because we can, but instead, “should we?”

Given the small scope I decided to start with, it took me just 2 days to get a version of this that I was comfortable sharing with others. Now it’s already out there, behind a feature flag for select folks to provide feedback on.

It was probably working in half a day, and the remainder of time spent on polish and testing. It’s truly remarkable what is possible now if you can build upon solid architecture and point your AI tools at good internal documentation.

Given that backdrop, I’m focusing on value delivered over time or effort spent. An ability to deliver something in 2 days does not necessarily reduce its value to the business owner seeking more clarity on their business.

If anything it highlights that there has never been a better time to experiment with building niche products that don’t scale. It’s a big part of my approach with Area 246 Labs.

If you’re interested in chatting more about this, reach out.

Learn More About Area 246 Labs

In the tech world, we are often conditioned to think of momentum as a motor. We are encouraged to grind and keep pushing until we hit an exit. As someone building solo, this feels dangerous for me given my prior post.

Motors require a constant, external stream of fuel. When the fuel, whether it is dopamine, novelty, or caffeine, runs out, the motor stalls. Getting a cold engine to turn over again can be harder than if you had just taken a more measured approach from the start.

The Battery Philosophy

Nowadays, I try to approach my work as if momentum is a battery. It has a specific set of characteristics that reflect the reality of solo building:

• Finite Capacity: You only have so much “high-voltage” focus in a day.

• Natural Decay: Energy leaks if it is not used intentionally.

• Cycle Life: Most importantly, a battery has a limit on how many times it can be recharged before it starts to fail.

If you “deep discharge” your battery to 0% every day, you are not being productive. You are permanently damaging the cells. This is what burnout looks like.

Avoiding the “Giant” Drain

Certain things act as a parasitic drain on your battery. One of the biggest is Comparisonitis, or in other words, looking at what established giants in your space are doing.

It is vital to remember who you are building for, not who you are competing against. Often, focusing on the user leads to a realization that there are many problems giants are now too big to solve. These gaps might not move the needle for a 500-person company, but they can still have big impacts for you and your customers. This is where I plan to spend my time with Area 246 and projects like Cari and Bajan Buddy.

Matching Task to Voltage

I also constantly remind myself that small progress is still progress. I try to match my task to the energy level I have rather than trying to “fake it until I make it.”

• On High-Voltage Days: I tackle the big features and complex architecture.

• On Low-Voltage Days: I do the small tasks like documentation, UI tweaks, or backlog grooming.

That is okay. It is similar to gym days where even showing up feels like a win. Small tasks also sometimes get me warmed up in a way that a large, looming task might not. It helps me avoid the friction of a “cold start.”

Be Back Next Month!

I am about to go on vacation for 3 weeks and I am explicitly not taking my computer. Why? Because there has probably not been a day this year that I have not been on it in some form. Time for a bit of an extended recharge!

Check out Area 246!

It’s 11am. I’m deep in something: a feature, a thread, a rabbit hole that felt important 20 minutes ago and probably still does. Then I hear it. The cat is meowing.

Not because of an emergency. Not because anything is wrong. He just wants to eat, and he doesn’t care what I’m in the middle of. Every day, at the same time, it’s a non-negotiable.

Left to my own devices, I could keep going all day. I know this about myself. The AI-assisted coding sessions that stretch late into the evening, the Instagram scroll that somehow becomes 45 minutes, I can’t always naturally stop. I optimize for momentum and call it productivity.

The cat is the most physical one, but he’s not the only thing pulling me out of loops I wouldn’t leave on my own.

In Europe, Instagram gives you a choice: share your data for personalized ads and doom scroll forever, or share less and get a forced unskippable break every few minutes. It’s clearly designed to make the second option feel like a punishment. I picked it anyway.

The break was supposed to be the friction. Instead it became the thing that snaps me out of a 20-minute scroll I didn’t realize I was in.

Claude Code does something similar. When you’re in a momentum streak, executing on a task and then suddenly you hit that rate limit. Usage resets in 2 hours. It stops you cold. But that forced pause causes me to do something else rather than continue staring at a coding editor all day.

And twice a week, at 5:30pm, I have Portuguese class. In person. I close the laptop and walk out the door whether I feel like it or not. I’m late more than I’d like to admit. I’m always glad I went.

Sometimes the constraint isn’t an interruption, it’s losing something you didn’t realize had its grip on you. My Apple Watch broke about two weeks ago. I went to get it fixed and was told it would take a month because there are no official Apple stores in Portugal.

My first instinct was to just buy a new one, and I actually did. I ordered it the same day and then I cancelled it before it even arrived. I decided to try to wait it out.

And even when I got the alert that the fix was ready, I had a genuine moment of: do I even need it back?

Some constraints are just the reality of decisions I’ve made around how I’ve chosen to live and what I’m interested in.

I’m not in Barbados full time. Cari is built for micro-businesses there, the vendors and small operators running everything out of a notebook and a WhatsApp thread. Initially that distance felt insurmountable. But it’s started me down the road of recognizing I don’t need to do everything myself. That’s a longer story for another post but this combination of constraints drove me towards adapting a different approach.

The same thing happens when you’re building.

Cari runs on WhatsApp. One of the most natural ways for a vendor to log something quickly is a voice note, you record a few seconds and you’re done, move on. It should be an obvious feature but I don’t have it yet and I considered that a limitation I had to address.

But nobody has actually asked for it, at least not yet, so I focused instead on what they were actually asking for, like an ability to record not just expenses but sales to customers. I was treating a missing feature as a constraint when the real constraint was the story I was telling myself about it.

Each one of these triggered something. A pause, a redirect, a better question, a person I wouldn’t have found otherwise. None of them felt good at the moment, but I guess that’s kind of the point?

I used to think sustainability was about discipline. Building the willpower to stop, to step away. And maybe for some people it is. But I can’t always naturally stop and I’m starting to make peace with that.

What I’ve gotten better at is noticing which constraints are worth keeping, even when I could remove them. The Apple Watch is back on my wrist but it feels different now. I could change the Instagram setting. I could push through the rate limit with a different tool. These boundaries are quietly doing valuable work I’d miss if they were gone. They work with the person I naturally am.

Get In Touch

I relied heavily on my engineering background to keep things on track when building from scratch. But for non-technical builders, it’s a whole different ball game: you’re figuring out what to build and how to build it.

Here are three traps to look out for. One can break your project, another will slow you down, and the third will cost you time and money.

Trap 1: Trusting and Not Verifying

This trap can break your project. AI tends to be a people pleaser: it will tell you everything worked, even when it didn’t really check to confirm that it did.

Why it’s a trap: AI sometimes skips over important verification steps, even if you tell it not to. I’ve seen it tell me that tests passed or tasks were completed, even when things were actually failing behind the scenes. When you’re building from scratch, you might not even know what “working” should look like, which makes it easy to overlook small, hidden issues.

How to avoid it:

• Make documentation a prerequisite. Before the AI implements ANYTHING, have it document the plan. After implementation, ask: “Does the documentation still match what you just built?”

• If you don’t have tests yet, ask the AI to generate them, using well known testing frameworks. Even better, have it write the tests first BEFORE implementation.

• Learn one simple command to run tests manually. It’s important to verify the changes it says are actually there and working as expected to avoid compounding problems making your app unstable.

Trap 2: Including More Than You Need

This trap will slow you down. When you add AI to existing projects, you can spot over-engineering because the AI’s additions feel out of place with your setup, but starting from scratch can make it harder to spot because so much is often happening at once.

Why it’s a trap: Over-engineering is often the result of unclear goals, leading you to implement unnecessary features. If things take off later, you can add scalability, but at this stage, you don’t need Netflix-level complexity. It’s easy to get caught up because something “seems like the right thing to do” or because the AI offers it. But these extra features slow you down and complicate your project.

How to avoid it:

• Break work into small, tracked pieces you understand. Instead of “build authentication,” try “create a login form” then “add password validation.” Have the AI create a task list upfront so you can reference it when you encounter limits or need to resume.

• Ask for the simplest solution that works. Always ask: “Is there an established tool or framework for this?” When established solutions exist, use them, they’ve been debugged by thousands of developers.

• Use version control. Ask the AI to automatically commit its changes when it has completed specific tasks, including clear messages about what changed. These checkpoints let you roll back specific changes without losing too much if something breaks.

Trap 3: Using One Model for Every Task

This trap can cost you time and money. Behind every request you make is a trained AI model trying to understand and execute what you need. It’s tempting to use the same model for everything, but not all models are suited for all tasks, which is why they vary in cost and capabilities.

Why it’s a trap: Lower-cost models often lack the reasoning power required for complex decisions, leading to inefficient code and poor planning that delays work. Meanwhile, using expensive models for routine tasks can drive up your costs without providing any real benefit. When building from scratch, it’s crucial to use the right model for the right task.

How to avoid it:

Think of it like a beehive: you need both a queen bee and worker bees.

• Queen bees (advanced models) should handle the big-picture tasks like planning and architecture, deciding what to build and how it should work (e.g., Claude Opus). Save the plan so both you and the AI can refer back to it.

• Worker bees (simpler models) can handle the implementation, writing the actual code (e.g., Claude Sonnet, GPT-3.5).

• Take breaks. Like human workers, worker bees can get stuck. AI can also get overwhelmed with too many requests. Resist the urge to keep pushing; sometimes taking a break will help you (and the AI) get back on track.

The Bottom Line

AI makes it easy to move fast, but moving fast comes with its own potential challenges. This is how you end up in the 80% trap: your prototype feels almost done, but foundational problems make the final 20% seem never ending.

Bonus: If you’re working with engineers instead of building everything yourself, here’s how to make that collaboration more effective.

If nothing else, make this habit stick: the next time you ask AI to implement something, have it fully document the plan first. This reference will guide both you and the AI moving forward.

Products are a collection of decisions made over time. Written plans help your future self understand the decisions you made, and why.

Subscribe now

Imagine you tell your AI agent to automatically buy that TV you have wanted for months as soon as it falls below $1,500 and to pay with your debit card.

You have expressed your intent, but unlike a traditional checkout experience there is still a great deal about the purchase that remains undefined. You do not yet know which merchant will fulfill the order, what the final charge amount will be, or exactly when the purchase will take place.

Traditional payments have typically been categorized into two buckets: Card-Present, where you’re physically at a point of sale, and Card-Not-Present, where payment data is captured remotely on the web or in an app. To me, this progression to agentic commerce seems like a brand new category: what Google, PayPal, Coinbase and others are calling Human-Not-Present1.

This payment type introduces three key uncertainties: which merchant you’ll transact with, when the transaction will execute, and what the exact amount will be. It assumes that you won’t need to be there to approve the final transaction at the moment it happens. We’re not there yet, though.

Even recurring subscription payments, which some might view as Human-Not-Present, feature more explicit user authorizations, providing a level of certainty absent from open-ended tokenized intent.

Consumers and B2B buyers are complex. There’s often a gap between what we intend, what we communicate, and what we actually do. For instance, you might tell your agent ‘buy the cheapest option,’ but actually prefer supporting a local business if the premium is minor. Or you state a price limit, but if your favorite color goes on sale just above that threshold, you’d want it anyway.

Purchases are about so much more than price. Logistical factors like cash flow timing, and personality factors like emotions and values can play a big role in us choosing what, when and how we buy, as well as the specific brands and merchants we want to support with our money.

This complexity is precisely why different merchants will react to agentic commerce in different ways. Companies like Walmart2, selling staples with relatively predictable demand and intent, are embracing it and it’s not hard to see why. Imagine shift and gig workers being able to automatically buy groceries the second their paycheck is deposited.

Or the potential for food waste reduction, where they could incentivize purchases of perishables with discounts right before they spoil, a sustainability win mutually beneficial for both the merchant and consumer.

On the other hand, for discretionary spend, like travel, where demand-based pricing already means we often pay different prices, I could see agentic commerce leading to mitigations to protect their bottom line. They might limit agent purchases to only their most restrictive fares, preventing AI agents from constantly rebooking whenever fares drop, and us pocketing the difference.

The irony? These same companies will likely embrace agents for trip planning and recommendations, which drive revenue, while attempting to block the transaction optimizations that could benefit consumers.

In the US, I’ve long held that we focus more on earning rewards from spending than returns from saving. Merchants will exploit this, using loyalty programs to encode brand preference into your intent profile before the agent starts shopping. This enables them to avoid commodification and makes loyalty a key battleground in the fight to shape our digital intent.

These loyalty programs themselves are likely to become more tokenized, enabling agents to auto-redeem rewards and optimize card selection, providing both expanded utility for users and new ways for merchants to keep you locked into their ecosystem.

The travel sector already demonstrates this strategy. Major hotel chains won’t provide you with loyalty perks like free breakfast or that room upgrade, unless you book via their desired channels. You could see more industries following this playbook.

But this automation at scale also introduces new challenges. As reported by the WSJ3, Apple Card’s synchronized billing, where over 12 million cardholders receive their bills simultaneously each month, creates a tradeoff: users love the predictability of the end-of-month due date4, but it creates concentrated spikes in demand for Goldman’s customer service teams and systems, causing problems not encountered with staggered due dates common elsewhere.

Now imagine a future where hundreds of millions of AI agents are all trying to execute transactions simultaneously in a similar bursty manner, whether buying concert tickets the moment they drop, or processing payroll-linked purchases every other Friday.

Even longer term, I can’t help but wonder if this shift could accelerate the drive towards tokenization of real world assets. As users get accustomed to digital representations of themselves, their preferences and their intent, why wouldn’t that also expand to the actual goods, services and assets they’re actually purchasing? From something as simple as a record of the expiration date of the milk in the fridge to something as complex as the deed to the house they live in.

This progression is happening, even if still nascent. As I explored in my recent post on zero-knowledge proofs, we urgently need to pair this future with privacy-preserving technologies that let us share this type of information, without exposing our complete intent profile to merchants and platforms.

There’s also the question of who actually wins in all this, and who gets left behind?

Get In Touch

This gap between what’s needed and what’s shared defines our relationship with data privacy and it’s about to become a much bigger problem. We overshare data points that comprise our identity every single day.

Digital identity has been a hot topic lately, with Apple recently making headlines for issuing digital credentials linked to your passport. A major benefit is selective sharing: instead of giving third parties access to every piece of personally identifiable information (PII) on your ID by default, you choose what to share, reducing the risk of sensitive data falling into the wrong hands.

What excites me about this is the potential for this to go even further. For example, a bar might request my date of birth or age when I tap my phone, but all they really get is confirmation that I’m over 18 (or 21 in the US). Nothing more.

With digital identity, the potential to provide abstracted validations like this without revealing the raw PII behind the scenes could be another leap forward in data privacy. The tech most often associated with this is called ZK, or “Zero-Knowledge” Proofs.

Apple and Google’s incumbent position as the devices and platforms running our digital lives gives them prime opportunity to educate consumers and take this kind of tech mainstream in a way that others cannot.

They are uniquely positioned to provide the necessary building blocks for 3rd parties to integrate, and just as importantly, have the leverage to be patient with regards to user adoption.

We’ve seen this play out before, where adoption starts really slowly and then accelerates once an inflection point has been hit. In mobile payments, Apple Pay launched in 2014, and its U.S. inflection point was COVID-19 six years later.

Passkey technology has been around since 2016, but the wave of data breaches in the 2020s pushed websites to adopt it, especially as most people still struggle to create strong, unique passwords. Still, outdated security questions like “First Concert Attended” or “First Pet’s Name” remain far too common for a generation born online.

With ZK tech, I can’t help but wonder if AI will be that inflection point. In a coming world where AI agents are making purchase decisions, running business workflows and taking other actions on our behalf, it becomes incredibly important that any data points and authorization necessary to execute those actions are available and confirmable.

This could open new avenues for exposing PII and sensitive data unless privacy-preserving solutions are built into AI tools, platforms, and partnerships. An option to “Hide My PII,” similar to Apple’s “Hide My Email” feature, which creates tokenized burner email addresses, would be incredibly appealing to me.

There is also another risk here though. The problem isn’t necessarily the tech, but more so what is deemed important or necessary to validate. By making it easier to request PII information, even if abstracted away from the specifics, we may see companies that don’t actually need this information requesting it in their apps and websites simply because they can.

OpenTable now tells restaurants whether you’re a ‘big spender’ or ‘often late’ before you even arrive for dinner. Your iPhone’s battery level can be used to track you across the web. A fitness app’s anonymized running data exposed secret military bases.

It’s going to be a wild few years (and decades) ahead in data privacy, especially when I think about all the various data points that will likely be required to deliver an outcome where autonomous agents can effectively act on our behalf, all day, every day.

And unlike handing your ID to a person in a store, you may not even know what data your AI agent is sharing, or with whom, unless privacy tech is built in from the start.

When thinking about Cari, my WhatsApp-based expense management tool for micro-businesses, the most notable tools in the space are products like Quickbooks. People know of Quickbooks even if they’ve never used it, but that actually isn’t actually my competition in this segment.

My competition costs less than $5, has no subscription, can be used in infinite ways without a learning curve, and a single one lasts months if not years. It’s a humble notebook.

It’s a surprisingly durable tool, pages can get dirty or tear, but it’s easy to simply throw into a bag and safer to leave out in the open by your side as you interact with customers all day, every day.

It is astounding to see the various “systems” folks have conceived to track their sales, appointments, inventory and expenses on the blank slate that is a paper notebook. It is truly the original Microsoft Excel.

Speaking of Excel, in some of my early experiments I explored converting some of these paper workflows to Excel (or Google Sheets, which is free), but the harsh truth was that many of these folks never used spreadsheets or quite simply didn’t want to. Some would keep using their paper notebook and then wait until I called to check-in to have me update the sheet for them.

With that in mind, Cari has always been about simplicity over glamourous features. I want users to get in, get out, and go about their day like normal. Logging an expense is as simple as dictating a voice note or a text.

In a perfect world, it’d be like electricity and the other utilities that we access multiple times every day without actively thinking about it. That utility today for many micro-businesses is a paper notebook.

Rather than directly competing with it, my focus is on chipping away at it and even working alongside it to deliver new benefits not possible with only paper.

Computer vision and AI can do a lot to build a bridge between existing paper based systems and software such that folks can finally gain automated insights and visibility into their business, even if they keep their paper notebook around.

Cari is a gambit on being there when the notebook falls short. I’m excited to explore more of that.

Learn More About Cari

Get In Touch

First up, what is Cari?

Cari’s goal is to provide a suite of business essentials for micro and small enterprises (MSEs) in emerging markets like my birthplace of Barbados.

While there’s no shortage of SMB tools like QuickBooks and Xero, they often feel overly complex to get started with, even for someone like me with an accounting degree.

Picture a street vendor selling food on weekends to make a bit of extra cash, or a tiny two-person business where one person takes orders and the other handles the work or delivery.

While my goal is to offer a suite of business essentials, I needed to start somewhere, and I chose expense management. I focused on this because, in conversations with MSEs, one theme kept coming up: they often have very little clarity on the overall health of their business.

Many rely on proxies like cash on hand at the end of the day or remaining supplies, but this doesn’t account for bulk purchases or other non-daily factors.

I built the product with extensibility in mind so I could quickly add new use cases in the future, such as appointments, which are predictive, whereas expenses are more historical.

How does it work?

It was important to make access as simple as possible and to go where MSEs already are. As such, Cari functions entirely in WhatsApp. There is no new app to download or website to login to, you simply message Cari to get started and go.

You can send it receipt images, manually log an expense by specifying the details, and request summaries and exports of your transaction history, all using a natural language conversational interface.

I then use a small and cost-effective LLM to convert that message into an API payload that can be processed against the backend, where the necessary data validations and permissions checking can be enforced.

Building out a product on a 3rd party platform where you have limited control over the UI is not without its challenges. My time at Buddy Media (now Salesforce Marketing Cloud) having to navigate Facebook’s ever changing platform policies meant this wasn’t my first exposure to this.

That being said, WhatsApp’s market penetration makes it incredibly attractive for this target segment, if you can get a bit creative with how you execute certain workflows in the product.

• Onboarding: Sign-up is done entirely via chat, so I built out a micro-KYC flow that users go through, including providing some basic information about their business. This helps me understand who is using the product, but also enables a potential recommendation engine down the line for improving their business.

• Currency & Locale: One benefit of WhatsApp is that your phone number does provide some signals which could be useful, because it ties you back to a country. With the country, I can identify the language and currency you likely want to use.

• Billing: Unfortunately, WhatsApp only supports integrated payments in Brazil and India, with no plans for broader rollout. To work around this, I built a flow that sends users to an external payment link, then redirects them back to their WhatsApp chat with Cari once payment is done. Webhooks help me track subscription updates and payment issues, so Cari quickly knows if your subscription lapses 😅.

• Exports: I believe in open data, so I added the option for users to export their transaction information anytime, whether for personal use or to share with a bookkeeper they likely already communicate with on WhatsApp. Given the sensitivity of this data, I implemented auto-expiring download links for these CSV exports.

How’d this go?

I could have moved faster

• My brain often naturally gravitates to platforms, rather than narrow use cases and my thesis in essence pushed me towards extendability rather than locking down a single use case.

• This inevitably slowed me down to start, although I am starting to wonder if the underlying platform components are now more valuable than any specific use case I could build out.

• That’s hardly a bad outcome, but potentially pushes me upmarket towards medium businesses looking for more customization rather than an out-of-the-box toolset for MSEs.

I could have learned less

• Cari is composed of DevOps tooling that I was not too familiar with when I got started: AWS Lambda Powertools, CloudFormation, Serverless Application Model, etc.

• This took so much more time than I expected to get setup, locally on my machine, and then in staging and production environments in the cloud.

• Individual components weren’t so bad, but building out a well oiled testing and deployment workflow wasn’t fun, especially as it’s sometimes difficult with AWS to fully develop and test locally.

• I learned a lot, but I also could have moved faster with something like a Heroku or Supabase, with potentially less vendor lock-in from AWS.

I should have considered using AI more quickly

• Is it weird that this is an AI-enabled product and I initially started building it without using AI at all?

• This slowed me down, no doubt about it, especially as a solo builder, but I also think this helped me develop a very strong understanding of the fundamentals and components underpinning Cari because so much of that core code I wrote from scratch.

• When I did finally start using AI to accelerate things, I found that knowledge incredibly helpful in how I devised my LLM prompts and instructions, because I could then quickly tell when things were about to go sideways and/or when it was suggesting changes that didn’t make sense given the context I had.

I had to approach testing differently

• As I alluded to earlier, Cari is split into 2 main buckets:

  1. The AI layer that assigns a user’s message to a specific workflow and the necessary API request payload.

  2. The backend API that actually has almost no AI magic, it’s a standard REST API and database interface.

• This is great as I can write and run tests against either bucket independently, and API testing is much easier than AI testing, so by enforcing validations at the API layer, I can more easily drive data consistency and integrity regardless of what the LLM comes back with (e.g. hallucinations, etc).

  • It also means I could build a full-featured app or website on top of the API rather quickly, particularly helpful if Meta’s WhatsApp policies change.

• This also freed me in how I test the AI components, because I can test a bunch of various messages and just confirm that they convert to the appropriate API payload without needing to always go end-to-end.

Why I did this

Being able to explore projects in depth like this is precisely why my focus is on fractional or part-time work to sustain myself. I readily acknowledge that sometimes it would be so much easier to just go find a full-time job somewhere, but I really enjoy this newfound space to explore during my prime working hours even if it means some opportunities just won’t make sense for me.

You can learn more about Cari here. If you want to chat more about Cari, or if you’re interested in comparing experiences and insights around building, feel free to reach out! I’d love to explore our different approaches, share challenges, and maybe even collaborate on future projects.

Get In Touch

P.S. If you’ve never seen Schitt’s Creek, it’s well worth the binge watch. My hubby and I have rewatched it COUNTLESS times.

One way I’ve loosely tracked the EUR/USD exchange rate is through my 5 EUR per month cell phone plan. It’s swung from as low as 5.22 USD to as high as 5.90 USD this year. A 13% currency swing in 10 months is substantial and can wipe out a year of typical stock market returns.

But beyond that volatility, there’s another issue: how prevalent dark patterns are in international payments, both in-person and online.

Jargon check: A dark pattern is a design practice used in websites, apps, or digital products that influences users to take actions they might not have intended. In payments, this can occur with Dynamic Currency Conversion (DCC) options.

Let’s start with in-person payments. Quite often when I use my foreign card, the card reader prompts me if I want to spend in EUR (local currency) or USD (my bank’s currency). Most folks know by now that choosing USD leads to an awful exchange rate and a 2-5% markup. As such, many folks default to saying NO to DCC.

Rather than accepting that initial NO selection, many payment readers here in Portugal will ask you the question again, but framed slightly differently such that a customer not paying close enough attention might blindly accept.

Merchants often receive a kickback whenever DCC is used and that has led to some questionable practices. Some merchants now state that it’s their policy that they MUST charge me in USD, with one recent restaurant claiming it was a tax that I needed to pay. They were visibly annoyed when I cancelled the transaction and used my Portuguese debit card instead.

Contactless payments have inadvertently enabled this behavior. Because most people just quickly tap the back of a payment terminal and never see the DCC prompts, they’re unknowingly leaving the currency choice up to the merchant. I’ve even had a merchant literally snatch a reader out of my hand and tap USD themselves in an uber-touristy part of town.

Online transactions shouldn’t have this issue right? After all, users are fully in control of what they choose during the checkout experience, but that’s not the case. Some websites make it seem like there is no choice. They default to USD and hide the currency option under opaque language or scare screens and messages that make you not want to choose them.

I would truly love an option to disable this functionality on an issuer or wallet level, but it’s unlikely to happen as the incremental revenue from this is likely in the billions annually if you consider overall cross-border payment volumes. Even Amex, which restricts DCC on its cards, anecdotally seems to just capture the full margin themselves with less competitive exchange rates than Visa or Mastercard.

This is not primarily a technology problem, but a matter of incentives skewed against consumers in cross-border commerce. Don’t even get me started on remittance payments to places like my birthplace of Barbados! So here’s my question: What would a truly consumer-aligned payment experience look like here? And what could cause enough pressure to make it happen?

• Almost every company feels like they NEED to do SOMETHING in AI

• More industry-specific, banks and fintechs WANT to do SOMETHING with stablecoins

This feels eerily similar to the mobile app rush of 2008, where every company suddenly felt like they had to release a mobile app, leading to some truly horrendous mobile experiences. Even now there are still many terrible examples, something I encounter most often when trying to book travel, which is ironic since it’s an industry built on mobility.

There’s also the mobile wallet rush of 2014 - 2020, which I had a front row seat at during my time at Braintree. Banks and payment networks launched a slew of me-too products as a protectionist measure against the then new kids on the block: Apple and Google Pay, and also to attempt to unseat PayPal.

In their rush to do something, many products were launched that drove little user interest and actually increased checkout friction, which is a big no-no for merchants trying to get users to complete their purchase as quickly as possible.

Doing something because you can is not enough anymore: the bar has been raised, which is great for consumers. The mobile wallets of today are better because of it. That bar applies to AI and stablecoins too. Nowadays, when I talk to prospects about integrating AI functionality into their business, I find it important to recenter them away from the flashy use cases that lack depth and towards proven use cases that deliver value sooner.

The chasm between prototyping a product and launching is an underappreciated challenge, especially given how quickly you can now get to that “80% done” milestone. But especially in areas like Fintech where the unhappy path is often the one that causes financial, reputational and regulatory risk, it’s even more important to get things right.

This might mean starting smaller to test the waters or looking internal first. I personally believe operations is one of (if not) the fastest paths to return on investment in AI today, with the added benefit of being able to explore in a controlled environment without directly impacting users.

It’s often an under-resourced part of the business where small wins can have outsized impacts. You can also safely assume that many internal folks are using AI unofficially anyway leading to no lack of ideas that could be worth exploring.

Switching briefly to stablecoins, there are already hundreds of stablecoins out there, and smart folks think there’ll be 1000+ (relevant xkcd comic below). The landscape is set to become more fragmented than prior cycles, though to some, that’s the whole point.

Ultimately users will make their choices in an increasingly crowded field, but for mainstream users who are used to complexity being abstracted away, this will be a challenge in the near term as it’s not just about what you hold, it’s also about what the person on the other side of the transaction is willing to accept based on the perceived risk profile, real world utility and other factors.

In traditional payments, users worry very little about whether their specific bank’s card will work for a transaction. That complexity is mostly abstracted away by Visa or Mastercard, and mainstream users (and merchants) will want that same simplicity for day-to-day payments. There are definitely contenders looking to become that new abstraction layer. It’ll be interesting to watch.

If you’re building in Fintech or AI and want to have a chat, reach out!

Get In Touch

While I do believe that finding a good fractional technical leader can be incredibly beneficial to long-term success, if you’re super early or budget constrained, I did want to provide 5 tips that enable you to level up with engineering until that time comes.

#1: T-Shirts: Not Just For Your Wardrobe

The challenge: If you're non-technical, it may oftentimes feel like a blind leap of faith when engineering tells you how long a particular task will take. This becomes even more challenging with hourly contractors whose incentives aren't aligned since they maximize income by maximizing hours.

What to do: Start with a t-shirt sizing system where you ask engineering to size work according to a scale of S, M, L, XL. This:

• Gives you visibility without the complexity of precise numbers

• Builds trust through understanding their reasoning, e.g. "this is medium because it is similar to the login feature we built last month"

• Creates a shared language over time and enables pattern tracking that helps your planning

#2: Sometimes a different garment is needed

The challenge: Even with t-shirt sizing, not everything goes to plan. Something estimated as Small can become Large once implementation begins, or perhaps what was built doesn't work as intended.

What to do: As a non-technical person, if you don’t have a technical advisor who can assist, I still encourage you to jump in on creative problem-solving and not to leave it solely to engineering to handle.

Your business perspective brings fresh eyes to the problem. You can help them step back and avoid sunk cost thinking by diving into whether the assumptions that underpinned the initial solution are even still valid. Is there an alternative idea you can verbalize that might be good enough for now? Sometimes you stumble onto something even better than originally intended.

#3: You need both a dressing room and a runway

The challenge: Prelaunch, it’s common to have a single environment for your application. However, once you have launched with real users, your engineering team needs a "safe" place to implement and test changes without impacting those hard-won users. “We’ll be careful” is not a strategy.

What to do: Have your technical team create a distinct test environment that accurately represents your application. Do this before launch, but even earlier if you can. It can be surprisingly difficult to recreate your application with its 3rd party tools, configs, APIs, etc. I’ve seen this delay launch dates when teams face unexpected issues getting what was already built ready for production.

Pair this with a simple deployment and rollback plan that documents how to get your application from the dressing room to the runway, as well how to return to the previous working version quickly if needed.

These plans are something I often help non-technical founders with, as it can be tricky to craft something simple and approachable that is also effective.

#4: I don’t know how to fasten this fancy belt

The challenge: I often hear from non-technical founders that they wished they had a better understanding of how their product actually works under the hood.

What to do: Modern AI tools make documentation faster and easier than ever to create and consume. Request that your engineering team start to generate easy-to-read documentation (system flows, third-party dependencies, etc) that show how things work.

Couple this with asking for unit and integration tests that validate functionality before users see them and you start to protect your product and users from the impacts of critical technical knowledge going out the door unexpectedly due to team churn.

Better yet, an incredibly powerful unlock for non-technical folks is the ability to couple this documentation with AI to generate a Q&A tool which they can query to help them make more informed product decisions and improve their communication with engineering.

#5: Completing the look

The challenge: You now have all the pieces: your sizing system (t-shirts), backup plans (alternative garments), environments (dressing room + runway), and documentation (the fancy belt), but how do you ensure everything works together seamlessly?

What to do: If you haven’t already, establish regular "outfit checks" (stand-ups) that tie everything together through consistent touchpoints. These create predictable communication patterns, catch issues before they escalate, and help maintain shared alignment on priorities.

The key is brevity and focus. This isn’t about micromanaging and should feel like an enabler for engineering rather than an additional burden. This approach is vastly superior to simply throwing tasks over the wall and waiting for results.

If establishing and maintaining these practices feels overwhelming, consider finding a fractional technical leader like myself, who can help implement and guide these processes until you're ready to bring on full-time technical leadership.

Book Free Intro Chat

The Chasm Between Prototype and Product

Prototypes are meant to be scrappy, as the faster you validate your product, the better. In the interest of speed, corners are cut, and especially for non-technical founders, you may not know what you don’t know, which means you don't know what to tell your dev team (whether real or AI).

This early phase is a bit like an extended hackathon: exciting, fast-moving, full of visible progress. You see features coming to life, early users signing up and maybe even some revenue flowing in. You’ve built something that technically works, it feels like you’re 80% of the way there. Maybe more.

But that final 20% is often where the extended work lives. It's handling edge cases when users do unexpected things, or addressing the random bug that actually isn’t that random and seriously impacting your users. It's this “unglamorous” work that separates a promising prototype full of your assumptions about how users will use your product, from a product built for how users actually do use it.

This post-validation moment is an inflection point, the stage where the right fractional partner can have outsized impact, especially if budget kept them out earlier.

From Frankenstein to Foundation

Most non-technical (and technical) founders start with what I call a "Frankenstein" product: a patchwork of different tools, manual processes and quick fixes stitched together. That's not just okay, it's often smart, as you get to focus on what makes you different instead of commoditized features.

But at some point, that Frankenstein product needs to evolve into a solid foundation for you and your users. Here’s how to tell if it’s time:

• You’re not sure which features are actually driving user value, or what metrics to be focusing on. Everything feels like a guess and the increasing complexity is making things worse.

• Your roadmap feels like a wishlist rather than a plan. You have no idea if estimates are accurate. When your developer says "that'll take two weeks," you have no context for whether that's reasonable. 

• Every update breaks something else. You fix the login issue, but now the payment system isn’t working. You’re stuck in debugging hell with little confidence in releasing new features or making changes to existing ones.

• All knowledge lives in one person's head. Your contract (or AI) developer is the only one who understands how anything works. There's little-to-no documentation, tests or structure. You're not sure what happens when code is "deployed," or where your data lives, or how to recover if something goes wrong or if your dev gets hit by a bus. This also makes hiring new devs even harder.

• You're making expensive mistakes. Using a technology stack nobody else knows, or paying $2,000 a month for something that should cost $200. The manual processes are taking too much time. This is more than just time and money, it’s a signal you need a partner.

The right fractional partner can help you bridge this gap systematically, transforming your patchwork solution and processes into the solid foundation your business needs to scale.

Flexible Ways to Partner

Typical engagements are two-staged:

• Stage 1 (2–4 weeks, depending on complexity): An audit of your current setup, followed by delivery of a product roadmap and prioritized recommendations that balance growth with stability.

• Stage 2 (Ongoing partnership): Continued collaboration with you and your development resources to support implementation of that roadmap.

What they won’t do is write all your code or be available 24/7 for every technical fire. Think strategic liaison, helping you scope, prioritize, de-risk, and sequence work so your team isn’t just building, but building the right things in the right order.

If an ongoing relationship doesn’t make sense for your stage or budget, consider doing just the stage 1 audit where you get the roadmap and recommendations, but then take full ownership over implementation. It’s a smaller upfront investment that can prevent costly mistakes, plus you can always decide to add ongoing support later.

Making the Call

The best time to bring on a fractional partner is usually 6-12 months before you think you’ll need a full-time resource. They can help you build the foundation and processes that will make that eventual full-time hire more successful.

If you're wrestling with these kinds of product and technology decisions and wondering whether fractional support makes sense for your stage and situation, let's chat about it. Sometimes a 30-minute conversation can save you months of headaches.

Book A Free Chat

BimPay aims to deliver instant, 24/7 payments that are mobile-first and don't require a traditional bank account. This is a genuine game-changer for financial inclusion in a country where in-person transactions still dominate, and across a region where financial services are often slow and burdened by high fees.

For consumers, you can pay for everything from groceries to utilities to rent using just your phone, just like how they manage their day-to-day life on WhatsApp. Having used instant payments regularly in Europe, I can attest to how transformative this becomes once embedded in everyday life.

For businesses, that cash-only vendor can now accept payments instantly with no fees or card processing costs, again using just their mobile phone. Secondary benefits include improved safety, reduced theft risks, and better record-keeping from handling less cash.

Building Beyond Payments: The Bigger Opportunity

Imagine a local university student who builds an app that helps fishermen not just accept BimPay payments, but also track sales patterns to offer automated savings for lean months. 

Or consider a Caribbean fintech like Onest using transaction data to provide income-based microloans with underwriting models trained on Caribbean spending patterns, finally giving informal workers long overdue access to financial services.

Similar to Singapore's current status as a powerhouse even with its small physical size, for Barbados, weekend hackathon projects could grow into Caribbean fintech success stories, but only if the infrastructure enables this level of innovation.

The Technical Foundation for Innovation

When Brazil launched Pix in 2020, instant transactions quickly became a baseline expectation for the (now 150 million) people using it, much like the immediacy of accessing electricity or water.

But Pix wasn’t just about speed. From day one, its design intentionally fostered a robust ecosystem and technical foundation to support innovation and exploration. For BimPay, this implies a need for:

• Open APIs that equip developers with the tools to build innovative solutions on top of BimPay, empowering entrepreneurs to create customized, locally relevant fintech applications that solve Caribbean-specific problems.

• Transparent Licensing & Regulatory Clarity that ensures non-bank players and startups can access the system through clear, fair requirements, leveling the playing field for businesses of all sizes, just as Brazil's Pix revolutionized payments by granting small vendors and startups the same access to payment infrastructure as incumbent banks.

• Test Environments, enabling developers to test prototypes and emerging payment solutions in realistic scenarios without using real funds or live data. This complements the existing Regulatory Sandbox, which allows companies to pilot with real users but carries higher setup costs and exposure risks. Simulation environments provide a crucial first step for early-stage experimentation before reaching the Regulatory Sandbox.

In smaller markets like Barbados, where startup resources and user bases are more limited, lowering barriers is especially critical. It accelerates validation cycles, reduces the cost of experimentation, and ensures that innovative ideas can be tested, refined, and brought to market without being stalled by infrastructure or regulatory complexity.

This environment could unlock a wave of new fintech apps, lending models, and loyalty systems, delivering outsized value to consumers, businesses, developers, and the broader economy.

Leading the Caribbean into the Future

Barbados is primed to quickly become a regional leader in financial infrastructure. In addition to BimPay, the island is slated to launch cashless payments for public transport this year as well.

Beyond just Barbados, the broader opportunity is with regional integration and interoperability. The region has a unique opportunity to leapfrog the interoperability challenges that have hindered financial infrastructure in other regions.

This push for connected payments is especially timely, given the region’s renewed ambitions for mobility of goods, services, and people driven by the CARICOM Single Market and Economy (CSME). Other initiatives like reducing cross-Caribbean roaming charges and improved regional travel options all contribute to that broader interconnected vision.

From Rails to Movement

The question isn't whether BimPay can move money, it's whether we will seize the opportunity to transform a payment system into a fintech ecosystem.

Balancing speed of execution with the thoughtfulness and intentionality that a slower pace requires isn't something I get right all the time, so here are 5 quick takeaways from my experience so far.

It can take over

• The accountability amplification is real - I've always been accountable for my work, but full ownership feels different. That 4am wake-up because something exciting popped into my head? Great for momentum, terrible for my rest and recovery goals.

• My laptop stays open - Sometimes it feels like I'm doing a solo hackathon instead of building a sustainable business. Reminding myself to close the laptop (or not open it) is an ongoing challenge.

• Maintaining work-life boundaries requires intentionality - Without external structure, it's easy to let work consume everything. I'm still figuring out how to maintain the balance I want.

It can get lonely

• Even introverted-extroverts need social stimulus - My home office setup lacks the coffee breaks and quick chit-chats that come from team work.

• It's not about collaboration, it's about presence - I don't necessarily need someone to solve problems with, but I do miss the social energy.

• Co-working with fellow solo founders is a good alternative - We mostly ignore each other while working independently, but there's something comforting about it.

Tests, Tests, Tests

• "Too early for tests" is not the way - It's tempting to say you'll write automated tests later, but at what point does it actually move to the top of the priority list? Why build a house without a basic foundation?

• Confidence correlates with execution speed - My ability to ship faster as a solo founder is tied to knowing I'm not constantly breaking things with a new feature.

• No peer review means tests are my safety net - There's something deeply satisfying about writing new code and validating with a single command that nothing (hopefully) has broken.

Documentation is like lighter fuel

• It forces better upfront thinking - Time spent documenting makes me do product and design pre-work instead of just "figuring it out as I go."

• I’m technical, but this is even more critical for non-technical founders - If you're working with external engineering resources, documentation gives you visibility into the "how" and "what" of their deliverables.

• This compounds in value over time - Like tests, documentation gets harder to bolt on after the fact and becomes exponentially more valuable as product complexity (and hopefully your user list) grows.

Non-human coworkers can add significant value

• Solo-founder doesn't mean solo-builder - I'm leveraging AI tools like a junior developer in the building process.

• As with real junior developers, the better their context (my documentation, tests, and project structure), the more valuable its output becomes, which drives a massive amplification of my capabilities.

• Without that context, like many human junior developers, you may end up with too narrowly scoped solutions or even worse at this stage, unnecessarily complex solutions.

All of this happens while I'm also doing Product Consulting and Advisory work, which feels like running two businesses where one requires constant client development. The mental context-switching between "building my thing" and "selling my skill set" deserves its own deep dive.

Speaking of which, I'm taking on new product consulting projects for Q3—helping clients think through strategy, roadmaps, and scaling decisions-particularly valuable if you're at that stage where you need experienced product thinking without a full-time hire. Drop me a line if that resonates.

Book A Chat With Me

I’m not even talking about building a full-fledged minimum viable product (MVP) – just getting to something tangible and interactive, that could spark the interest of potential cofounders, investors or clients on the potential value.

Technical folks, like me, have long had an advantage here, because we often have the ability to prototype our ideas without the costs that come from hiring a third party to do the work. [Sidebar #1: this can also be to our detriment and lead to us not validating our ideas early enough.]

Things have shifted dramatically over the past few years though. It is now possible for anyone out there to convert an idea into an interactive proof-of-concept (POC) with just a paragraph of text typed into an AI platform.

Crafting that paragraph is still a bit of an art, with its own quirks, but this will likely unlock a torrent of new value delivered to both first-time founders and their customers as the barriers to entry fall.

To be clear, a POC is not an MVP. Technical resources are still necessary to accomplish the V in MVP for a product with any level of complexity. Real users will require you to have a real understanding of what’s happening behind the scenes in order to sustainably scale and support them, but AI tools are enough to kickstart the P in POC. 

After all, a picture is worth a thousand words and anything that quickly enables validation and can drive interest in an idea can justify the additional effort needed to build an actual MVP from a POC. 

This means some great ideas that would never have been built before will likely come to light, creating real outsized value. At the same time, there is also likely to be a lot of low value boilerplate products that will arrive like a tsunami of noise.

We’ve seen this play out in prior cycles, e.g. the mobile app stores for iOS and Android now have over 1.5 Million apps competing for our attention, even as most folks use less than 30.

[Sidebar #2: WhatsApp had 50 employees when it was bought for $19B, Instagram just 13 when it was bought for $1B. It is both exciting and terrifying to think about what is possible today given the tools we have now.]

Even so, just like mobile apps are not by themselves a product (unless you’re Apple or Google), for most folks neither is AI (unless you’re OpenAI, Google or Anthropic).

I am very excited by what I have seen some non-technical founders deliver, including some consulting clients that I have worked with to get them from early concepts to full-fledged products with live users, integrated payments, forward looking roadmaps and ramp plans.

If there’s an idea that’s been living rent free in your head for some time, now’s the time to start exploring. The barriers have never been lower, especially for those historically underrepresented as founders (women, people of color, LGBTQ+, etc).

I am now myself a founder, building MicroSaaS products, self-funded from income I make partnering with startups and non-technical founders looking for product consulting support. If you’re interested in working with me, book a free discovery session here.

Book Free Discovery Session

Leave a comment

Micro, small and medium enterprises (MSMEs) have long been an interest area of mine. My entire family is full of people with hobbies that they’ve managed to turn into a side business and even full-time gigs. If anything, I’ve been a bit of an outlier, so far…

The quiet setting enabled me to ask him some questions about how he runs his business that might have been awkward to do at other times.

After a few probing questions about his day-to-day, my barber looked at me funnily and said: “You’re not looking to open a barbershop are you? If you are, I’m happy to help as long as it’s not too close to here.”

It was then that I explained to him, that I am doing some user research on MSMEs for a potential project, and that rather than trying to compete with him, it might be worse, I just might try to sell him something 😆.

I left that day with a lot more than a haircut. I hadn’t been planning to interview him, so as we wrapped up, I asked him if I could pay him for a 90 minute session in a few weeks, so that I could include a more structured 45 minute Q&A, which he happily agreed to.

Two weeks later, after my haircut, we went to a nearby cafe. As he answered my questions, I furiously scribbled down learnings and other details, and tackled uncomfortable but critical questions like willingness to pay.

At the end of the session, I went to pay him for this additional time and he flat-out refused my cash. The most I could get him to accept was me paying for the Diet Coke he consumed during our chat.

Doing user research is not a new thing to me, I’ve done a lot of it during my career as a PM, but in typical “do as I say, not as I do” fashion, it’s been less of a focus when hacking on my own stuff. In the past, my side projects were not really focused on external impact, but self-validation.

I relished having an idea and figuring out how to build it on my own, most likely learning some new tech along the way that I could add to my skill list. That itself used to be enough. 

I’ve almost certainly let a few decent ideas die this way, as “self-validation” is also another way of saying that I built in private. Most of my side projects never even got something as quick and simple as a social media post. It’s been almost 14 years since I built my first iOS app, for iOS 4 (not a typo)!

Over the past 16 months, I’ve slowly been pushing myself to do more things publicly, e.g. this very newsletter you’re reading. However, that random conversation with my barber was a blip in my typical perfectionist approach to going public.

The lack of pre-planning made me quite uncomfortable when I first started asking him questions on the fly, but the extremely casual approach to our conversation likely meant that I received more authentic responses than if I had started with a more formal interview.

My chat with my barber ended with “If you build this, I’ll sign up right away” and all that validation point cost me was a little courage to put my still half-baked idea out there. Oh, and that 1 Diet Coke.

In my final 2024 post, I reflected on the challenges I have faced navigating financial systems across countries.

The experience is not all bad, so as promised, here is a follow-up post where I share 3 of the unexpected arbitrage and learning opportunities that have also resulted.

But first, Happy US and Canadian Black History Month!

1. Arbitraging Discrepancies Across Borders

Loyalty points programs, like most things, are not immune to inflation. In the US, the cost of redeeming rewards have risen dramatically, with flights and hotels, especially on the higher end now costing 4-5x more in points than a decade ago. 

What is somewhat different however, is that unlike salaries, which have struggled to keep up (e.g. the federal minimum wage was last increased in 2009!), earning rates for loyalty points in the US have climbed drastically as well, even if not fully keeping pace with redemption rates. 

This situation is somewhat unique to the US, driven by the fact that US airlines in particular often make most of their profits (if any) by selling miles to banks, rather than from us buying seats, and banks are happy to pay up given the levels of consumer engagement they drive.

As someone who is still earning loyalty points at the increased US rates, but now redeeming them for travel that often does not touch the US, I’ve unlocked exceptional travel opportunities. Outside the US, there’s much less competition for rewards redemptions, often leading to much lower costs for flights and hotels.

For example, a basic NYC hotel room can cost the same points as a pool villa in Asia. A domestic US flight on Delta can often cost more than a 12-hour flight between Europe and Asia. 

This isn’t even just a loyalty hack. For example, Amazon Germany and Amazon Spain can often have different prices for the exact same product, both shipping to Portugal given they’re all EU members. Something not really discoverable unless you’ve had the joy of navigating global companies across countries.

2. MacGyvering solutions when systems fail

For those unfamiliar, MacGyvering means “making or repairing something in an inventive, improvised way, with whatever is at hand.”

When I encounter a "computer says no" situation, I dive into troubleshooting, often leading to creative solutions. For example, when a local website didn’t accept my international phone number, I examined the source code and tried different formats until the SMS came through.

Another example involved my monthly transit service. Every month, online top-ups failed with my international cards. I would try multiple cards and they all failed. One day, after the first failure, rather than try another card, I simply re-used the same card details, and it worked on the second try. 

Nowadays, I just expect the first failure, and simply retry the same card. It has consistently worked, avoiding the need to go in-person each month. A bit annoying, but (currently) acceptable workaround.

3. Leveraging my newfound flexibility to do things my way

In 2015, I found a $495 roundtrip Business Class fare on American Airlines to Beijing, available for just a few hours before it was pulled. My hubby had concerns about booking a trip with short notice, but we compromised by making it a super short trip (48 hours!). It was a fun experience, and perhaps a story for another post.

Looking back, I’d likely handle it differently now—perhaps staying longer while he returned to work, or even going without him, but my interest in solo travel didn’t truly emerge until 2019.

Now, with career independence, I love the flexibility it has enabled. My mornings can be dedicated to personal passions and my burgeoning “solo-trepeneurship” efforts, while afternoons are focused on consulting and advisory work.

I can’t say I’ve figured everything out yet, but I’d rather bet on myself, rather than going back to career autopilot anytime soon.

If you’d like to hear more about anything discussed here, let me know in the comments or get in touch.

The more that I put down new roots here, the more it has exposed the challenges in retaining my established roots elsewhere in the world, not just personally, or even professionally, but most frustratingly, financially.

Many banks take a lowest common denominator approach, solving for the biggest pool of people as simply as possible, unless you’re crazy rich. P.S. I’m not crazy rich.

Given that, the second I add foreign to my equation, you encounter a lot of “computer says no.”

I’m not that surprised to be honest, given banking and its risk-on/off oscillations over the last few decades. I’m likely dating myself here, but as a former Lehman Brothers intern before its 2008 demise, I’ve been through at least 3 financial crises, if we count the more recent crypto (FTX) and regional banking (SVB, First Republic) collapses as such. Not to mention the various sanctions, KYC and anti-money laundering lapses that have irked regulators worldwide.

🛑 Computer says no, with little recourse

My lack of surprise doesn’t make it any less challenging, as it often strikes at inopportune or random times, and solutions recommended by many banks often lack credibility or utility. Some recent examples:

• Upon replacing my phone, one bank prevented me from re-adding my existing card to Apple Pay until I returned to the USA. There was no other way to re-verify myself, even with an agent, which in situations where your physical card is lost, stolen or damaged removes a backup option when most needed abroad

• Many critical websites, including the provider that we use for our mortgage(!) won’t even load outside of the USA without a VPN. I can’t exactly just change banks or use a backup here

• I had one bank routinely block transactions on my card daily. I called every day, reminded them I’m abroad, they would say ok, you’re all set going forward, and then like clockwork it would be blocked again the next day. This carried on for weeks, and then the false positives magically stopped

• Many local websites around the globe will just fail to accept US credit cards because many major US banks have poor support for verifying card transactions in real-time via two-factor authentication/SCA/3DS. Sometimes merchants will also simply block all foreign cards because most of their customers are local.

We have some basic interoperability, but I crave true mobility 💸

As someone who has lived in the Caribbean, USA, Asia and now Europe, I am constantly reminded that financial infrastructure is not global, just somewhat interoperable. True global mobility is lacking in financial services.

This is expected to some degree, as regulations and their associated obligations/penalties are generally country/region specific, but it doesn’t make it any less frustrating. 

If you’re primarily a tourist, interoperability (mostly) has your back. Looking to send remittances back home? Interoperability has got you there too, but often at an extortionate cost. 

This is not limited to legacy banks. Even Apple, which pulled a magical hat trick this year by shipping me a replacement iPhone from the US to Europe for free in just 3 days over a weekend no less, still has digital servicing issues.

For example, I now have 3 Apple iCloud accounts (US, UK, PT) given how some apps and features are geo-restricted. It’s like Pokemon, gotta catch em all I guess? My digital life doesn’t end in one place just because I am now spending time somewhere else.

This isn’t even just a foreign issue; why must changing a job even in the same city often require me to manually set up direct deposit again, set up my retirement contributions again, and manually verify that I’m not inadvertently hitting annual limits? Likely some combo of slow regulation, legacy tech and misaligned incentives on Open Banking.

Could this get better? ⚒️

Unfortunately this is not a “ta-da” kind of problem. It’s a hornet’s nest of multiple problems spanning different financial products, regulations, currencies, customer profiles (this post doesn’t even consider businesses and their challenges), etc.

That being said, Revolut, Wise and other pan-European players seem to be on a mission to build the world’s first global “banks” for the non-wealthy, mass consumer segment. I’m cheering them on, as they leverage the rise of digital nomads like myself and the increasing popularity of remote work to great success.

One of my former employers, PayPal could be interesting given recent moves, and their existing prominence in checkout worldwide. Tip: They can often be the only way to pay when you lack a local billing address or payment method.

Amex has some offerings that allow you to transfer your credit history across markets, often one of the most challenging aspects of establishing yourself in a new country. That being said, good luck trying to use an Amex card in Portugal, but still very handy in other markets like Australia, Mexico, Singapore or the UK.

It’s not all bad

Finally, while it is sometimes frustrating as a user waiting for this to all play out, especially as someone who works in payments, identity and compliance, and being able to envision what’s actually happening behind the scenes, it’s not all negative.

There are definitely ways to capture outsized value as an outlier that has benefited me previously in Asia and now Europe. I’ll share some of those in an upcoming post.

If you’re working in this space in some form or have similar experiences, tips or advice, I’d love to catch up. Get in touch with me here.

Does anyone actually like interacting with their government to access key services or manage obligations like taxes? 

If we think of governments like companies, they would be some of the world’s biggest: literally everyone is their customer in some way.

That scale comes with an incredibly varied product suite, with some services accessed frequently, while others are super niche. The dense and cascading regulations that dictate everything also lead to high levels of complexity.

It’s not surprising that most users find it unsatisfying. This sentiment persists even for the world’s biggest governments, who have the scale to justify larger investments in improved service delivery. 

This is even more acute for a small government like my birthplace of Barbados, which requires many of the exact same services as larger economies, but spread over a population of just over 275K people, maybe 750K if you count tourists and other visitors.

Even if we can’t turn all government interactions into a trip to the beach, how do we make it more like an annual checkup than a root canal? It probably won’t surprise you that I think Generative AI has a huge role to play. 

Yes, that same Generative AI tech that I also categorized as a moody teenager, and previously ranted about its dangers and biases. 

Those things are all still there, but it also has the potential to modernize government services and adapt to changes more quickly than traditional GovTech solutions, especially in smaller economies and/or on the hyperlocal government level. 

Imagine the following…

Quick answers to complex and nuanced questions

• You as a person or entity are nuanced, with your query requiring precise answers

• Government policies1 can be very technical in their quest to be comprehensive

• This intersection often requires reliance on government workers (and accountants and lawyers) to get accurate answers

• This takes time because they too are human and may need to do research

• Imagine empowering these workers with a resource they can query in natural language with your specific situation. It then produces answers referenced from those complex government policies in seconds instead of minutes or hours

• To reduce risks, the worker can leverage the provided references (and other sources) to quickly confirm everything before responding to the citizen-customer

Now imagine that directly available 24/7 via WhatsApp

• Given the current risks of inaccuracies (which hopefully subside over time), this is riskier than an internal-facing tool

• But, sometimes it’s not just about the answer. Simply discovering who to contact across a complex hierarchy of departments can have value

• In that scenario, a citizen-facing router or directory can be impactful with limited downside risk. Examples:

  • Where can I find more information about registering a business?

  • What website should I use to make a passport appointment?

• As the technology improves you could then empower citizens further, enabling them to directly query their government and receive responses in natural language

• The fact that this can happen via channels they use every day, like WhatsApp, with no prior training, is a game changer. More examples:

  • What is required to register a birth or death?

  • What are the requirements for a passport?

  • What does this line item on my tax return mean?

  • What loan or grant programs are available to small businesses?

Not Just For Citizens

As I alluded to earlier, governments also serve more than their citizens or local businesses; there are tourists, students, digital nomads and other foreign workers who are even less versed in interacting with a government than its citizens. 

As someone who has “imported” his cat into 4 different countries in the past 18 months, I have felt this pain.

Solutions here could make early-mover destinations more attractive and raise the prominence of a market at a time where competition is fierce for tourism dollars, foreign talent and overseas investment. 

Enabling Balanced Regulation

Finally, much has also been said about the need for Responsible Generative AI given its far reaching impacts on society. Sound regulation could be better informed if governments had direct experience with these tools and their nuances, via prototypes or pilots targeted at specific use cases. 

Unsurprisingly, the Singapore government is already experimenting heavily with AI across education, tourism and other government functions.

Let’s Talk

Are you looking to work in this area, or already working in it? Heck, if the intersection of GovTech and Gen AI is just plain interesting to you, please reach out. I’d love to have a chat as I see multiple opportunities ahead.

Contact Me

Applying some structure and loose goals were critical to my continued development and progression, but in a different way.

Note: while I mention setting “goals” above, in reality I defined themes that guided me, rather than explicit deliverables. I do find myself mapping short-term goals to these themes to understand if they fit my still evolving north star.

Theme 1: Be more creative and learn non-technical things

So much of my development since I became an adult has been focused on professional pursuits. I wanted to expand my focus to some areas that would challenge me differently.

Photography

• I take a lot of photos: 6K in 2021, 10K in 2022 and somehow, 25K during 2023

• Given that level of increasing commitment, I wanted to get better at adapting to various shooting environments, including night photography, astrophotography, landscapes, etc

• I will admit, some of that rapidly increasing photo count is due to trial and error, but I’ve also learned that editing is half the job, and can help you recover questionable outputs, which might be a life learning as well 🙂

Language learning

• I never understood the privilege of having English as my native tongue, until I traveled to far flung places in Asia, South America and Europe, where it was the de-facto backup language

• That being said, my new home is Portugal and I want to fully embrace all that means, so I’m 5 months deep into learning Portuguese!

Theme 2: Slow progress is still progress

Even as I was slowing down in other ways, one area that I stubbornly found myself still doing too much too soon was physical fitness.

CrossFit

• I am a man of routine, and I started CrossFit thinking it would enable me to maintain consistency in working out, even as I moved between Barbados, Manchester and Lisbon

• That being said, CrossFit kicked (and continues to kick) my butt – sometimes my weights felt comical compared to others around, but it also made me focus on my own progression as the measure of growth, which was helpful in other contexts

Hiking & Biking

• Hiking is now one of my favorite things. The amount of random weekdays that I found myself up a mountain, or along a lake, or freezing my butt off on a snow capped hillside were substantial, and also helped me take on my fear of heights

• I also FINALLY learned to ride a bike y’all! Although it’s not as useful in hilly Lisbon

Theme 3: Get back to learning by building

Earlier in my career, whenever I wanted to learn more about some trending tech, I built something like an app or website with it. As I’ve progressed in my career, I’ve found that most of my recent exposure to emerging tech has felt more theory driven than practical.

Generative AI

• I consumed a lot of Generative AI content, and even earned a Prompt Engineering certificate, but it was only when I tried to build something end-to-end with it, that I truly understood its potential and its limits

Blockchain & Smart Contracts

• I’m currently mid-flight on a Ethereum Blockchain Smart Contracts course, and I’m really pumped to continue to dive deeper from a more technical perspective

Infra

• Building is more than just ideating, scoping and developing, you also need to figure out what infrastructure to use to deploy it, which can be tricky, expensive, a security risk, etc

• I dove headfirst into building “serverless” applications with Amazon Web Services Lambda, as there’s nothing quite like taking something that’s been running on your laptop, and getting it live in the “real” world

I’m certain my themes will evolve over time, as will the goals that they inspire, so I’ll continue to periodically share updates.

Até a próxima! 🇵🇹