Building Products in Morocco With AI

Every tool we use in Morocco was built somewhere else.

The CRMs, the marketing platforms, the payment processors, the chatbot solutions. Built in San Francisco or Amsterdam or Paris, for companies operating in markets with different infrastructure, different customer behavior, different languages, different payment methods. Then exported to us, with varying degrees of adaptation.

Sometimes it works fine. Sometimes it works fine enough. And sometimes the gap between "what this tool was built for" and "how business actually works here" is wide enough that you spend more energy working around the tool than using it.

I spent years in the second and third category. Then I started building instead.

What "Moroccan market" actually means for tech

When I explain to foreign colleagues why we built our own WhatsApp platform, or our own chatbot infrastructure, or our own image generation workflow, the first question is usually: "Why not just use [existing solution]?"

The answer requires a bit of context.

WhatsApp in Morocco isn't a messaging app. It's the primary business communication channel. Your clients will not use your ticketing system. They will not open your client portal. They will send you a WhatsApp. That's not a behavior you can train out of them; it's deeply embedded in how business relationships work here. A CRM that doesn't understand WhatsApp centrality is a CRM that works around the most important channel.

Payment in dirhams, without forex complications, without international card requirements: this matters more than it seems. A SaaS product billed in dollars to a Moroccan card is a friction point significant enough to lose deals. The businesses that operate in this market either accept dirhams or they leave a segment of it unaddressed.

Bilingualism (French and Arabic, often Darija specifically) shapes everything from interface design to AI model fine-tuning to customer support scripts. A chatbot trained on French and English that encounters Darija produces outputs that range from confused to actively wrong. The local context has to be built in, not bolted on.

These aren't edge cases. They're structural features of the market.

What AI changed about building locally

Before, building for this market meant long development cycles, significant cost, and the constant trade-off between building something right and building something now. A custom CRM designed around WhatsApp-first communication was a six-month project minimum. A local chatbot solution that actually understood the Moroccan customer context (the language, the industry specifics, the cultural communication norms) was a research project before it was a product.

AI changed the build timeline. Not by automating the thinking. The market understanding, the product decisions, the architecture choices are still human work. But by compressing the distance between decision and implementation. What used to take a development team six weeks now takes one or two, at the prototype level. What used to require external hire now requires better prompting.

Risala, our WhatsApp CRM, took weeks to go from concept to working internal tool. nBot, our chatbot platform, was built with AI assistance and now serves over 250 businesses. Not because we had a large engineering team. Because we had a clear understanding of the market and the tools to build quickly against that understanding.

The advantage of building for yourself first

Every product we've built externally started as something we needed internally.

Risala started because my own WhatsApp inbox was unmanageable. Suwar started because BannerBear didn't do what we needed. Barq started because we were paying for Mailchimp features we didn't use and missing the ones we needed.

This is not a coincidence. When you build something to solve your own problem in your own market, you build it correctly from the start. The assumptions are right. The edge cases you handle first are the real edge cases, not the theoretical ones. The UX reflects how you actually work, not how a product manager in another country imagined you might work.

The best products in any market come from people who deeply understand that market, not because they've researched it, but because they live it. AI makes it possible for those people to build, even when they're not professional developers. That's the real shift.

What still needs to be built

I'm not going to pretend the landscape is complete.

There are still significant gaps in what's available to Moroccan businesses, not in AI capability, which is global, but in AI application, which requires local context. A language model tuned on Darija, trained to understand the specific communication norms of Moroccan commercial relationships, is genuinely different from one trained on French and Arabic classical text. The capability exists. The application, locally, is still early.

The infrastructure for AI-powered financial services (credit scoring, fraud detection, insurance pricing) that reflects Moroccan data and Moroccan behavior rather than European or American proxies. The automation tooling for sectors that are significant here (agribusiness, logistics, construction) that speak to the specific operational realities of those industries in this country.

These things are coming. The question is whether they'll be built by people who understand the market or imported from somewhere else and adapted.

I know which I prefer.

The point of building in Morocco

There's something I believe that might sound idealistic but that I've watched prove itself repeatedly: the best technology for any market gets built in that market. Not imported and adapted: built from inside.

Not because foreign tools are bad. Because the understanding that comes from operating inside a context, from feeling the friction yourself, from building for a problem you live rather than one you've observed, produces something different. Something more right.

We've built things here that don't exist elsewhere, for problems that don't present the same way elsewhere. That's not a disadvantage of the market; it's an opportunity. If you understand it, and you have the tools to build quickly, the playing field is wide open.

That's where I'm standing. The field is still mostly empty.