Banks Spend $700bn a Year on Technology and Still Cannot Deploy AI. Stitch Is Building the Layer That Changes That.
When Andreessen Horowitz made its first investment in the Gulf Cooperation Council this month, it didn't go to a consumer app, a neobank, or a payments unicorn in the making. It went to a three-year-old infrastructure company operating in what is arguably the least glamorous corner of fintech: the plumbing.
The $25m Series A the firm led in Stitch, bringing the Saudi startup's total raised to $35m, is a quiet but significant moment for the region's technology ecosystem and a statement about where one of Silicon Valley's most closely watched firms believes the next decade of financial services will be built.
For Mohamed Oueida, Founder and CEO of Stitch, the validation carries weight but not surprise. “Financial institutions globally run on fragmented, legacy infrastructure that should have been left behind 20 years ago,” he said at the time of the announcement. “Now every institution wants to adopt AI, but AI on top of broken infrastructure is a dead end.” It is the kind of sentence that doubles as a founding thesis and a sales pitch. In Oueida’s case, the two were never really separate.
The Problem Nobody Was Solving
Before Stitch, Oueida ran a payment terminal business in Saudi Arabia, notable in itself as the first non-banking company to offer payment products in the kingdom. On the surface, the product was about merchants. In practice, it became a lens into something more foundational. Every problem a merchant had could be traced upstream to the banks themselves. The infrastructure that financial institutions depended on was not a series of isolated technical failings; it was a structural condition. Nobody was doing anything about it.
That observation became the organising principle of Stitch, which Oueida founded in 2022. The company’s product is the nervous system of a financial institution: a single cloud-native platform that replaces the hundreds of disconnected systems governing everything from account opening and KYC to card issuance, ledger management, and payment processing. A typical bank of meaningful scale runs 400 or more such systems. Larger institutions operate more than a thousand. Each holds data that is, in practice, inaccessible to the others. Deploying AI on top of that architecture does not produce intelligent systems; it produces constrained ones. The infrastructure gap must be closed before the AI ambition can begin.
Infrastructure Debt Has a Price, and Financial Services Is Paying It
The financial services sector has spent upwards of $1tn on digital transformation over the past three years. Most of it has not solved the underlying problem. Banks globally allocate an estimated $700bn a year to technology, yet launching a new product still takes years and upgrading a core system still risks bringing operations to a halt. The core systems running the majority of regional institutions were architected decades ago, designed for batch processing and siloed operations. Cloud migration, where it has happened, has often been a lift-and-shift exercise: moving old architecture to new hosting without restructuring the underlying data model.
In its investment announcement, a16z put the consequence plainly: “financial institutions are sitting on decades of infrastructure debt. This debt has become the single biggest obstacle to AI adoption. Intelligent agents cannot run on systems without clean APIs and data.”
The problem is not a niche concern. It is what keeps fraud detection from working at scale, credit decisioning from being real-time, and AI models from having anything coherent to work with. A bank running 400 systems, each holding a fragment of customer and transaction data that cannot communicate with the others, cannot deploy AI that works in production.
Oueida’s assessment of where most institutions in the region actually stand on AI adoption is deliberately uncomfortable. It is, he says, largely non-existent in any functional sense. Most large institutions have a proof of concept running somewhere. Extracting real operational value from it is a different matter entirely. The response to that condition cannot be to add another product layer on top of a broken foundation. It has to be to replace the foundation. That is what Stitch was built to do, starting at the database level, designed to be AI-native from day one.
Building the Team From the Ground Up
The founding team was assembled through a largely outbound approach: Oueida reaching out to prospective hires directly, often through LinkedIn chains that led somewhere unexpected. What united the group was not a shared employer or a common institutional pedigree, but a shared frustration. “Every single person is fed up with the way things are,” he said in a recent interview. Veterans of banking and payments, most of them drawn from institutions including FIS, Barclays, Santander, NPCI, and Azentio, who had spent long enough inside those systems to know that incremental improvement was not going to be enough.
In its investment thesis, a16z described Oueida as having “the rare combination of relationships, sales instinct, and grit to win in a market where trust and distribution matter as much as product.” That combination was tested early and often.
The First Clients Were Never Going to Come Easily
Selling into banking and financial services falls into a different category of difficulty than enterprise sales generally. Regulated institutions move slowly by design. Every new vendor relationship involves compliance reviews, security assessments, and, in the Saudi market, engagement with one of the most demanding regulatory environments in the region. For a startup with no track record, the door is not merely difficult to open. It is, for most, closed.
Stitch’s initial instinct was to find a way around that barrier. Before targeting banks directly, the company pursued embedded finance opportunities: non-traditional operators seeking to add financial products. The logic was defensible. The traction was eventually not sufficient. If the thesis was ever going to be tested at the scale it required, the team accepted they would have to go directly after the banks, absorb the slow sales cycles, build compliance credibility certification by certification, and wait.
The first real customers came through a combination of deep pain and, as Oueida puts it, people willing to take a genuine bet. Amodi Exchange, the oldest exchange house in Saudi Arabia, was running on a system implemented before the 21st century. Lulu Exchange in the UAE had spent years trying to launch a product through a succession of technology partners but had not succeeded. Neither was looking for another vendor. Both were looking for a way out of a situation that had become unmanageable.
What made those deals close was not primarily the product, though it had to withstand intense scrutiny. It was the nature of the relationship Stitch offered. Most financial institutions are deeply frustrated with their existing technology partners: over-promised outcomes, under-delivered systems, and no meaningful accountability after the contract is signed. Stitch committed to genuine partnership, absorbing the problem alongside the customer rather than delivering a system and stepping back. Both exchanges signed as paying customers from the first contract. There was no design partner arrangement, no free access in exchange for product feedback. “It was very painful,” Oueida said. “But in hindsight, a lot of people tell us it was a brilliant call.”
From those two customers, the credibility compounded. Each subsequent deal became marginally less difficult to close. The compliance and security certifications required to operate in Saudi Arabia, some of the most rigorous in the region, became a competitive advantage when Stitch expanded into other markets. The question of security, which immediately arises in any conversation with a regulated institution, had already been answered.
Why a16z Made the GCC Its First
The Andreessen Horowitz investment did not begin with a formal pitch. By Oueida’s account, the relationship developed gradually through the tight-knit global community surrounding banking infrastructure, a space where the number of people building at this layer is remarkably small. Jason Gardner, Founder of Marqeta, had participated in Stitch’s seed round and moved in the same circles. Alex Rampell, a General Partner at Andreessen Horowitz with a long background in fintech, was already closely tracking the infrastructure problem.
The conversations started casually: updates on what Stitch was building, nothing structured. They accumulated over a series of meetings before the company decided to raise a Series A. When asked what clinched it, Oueida pointed not to the metrics but to the team itself, a group of domain veterans choosing to build something difficult in a market most valley investors had not yet taken seriously. “They really just got it,” he said.
Alex Rampell, General Partner at Andreessen Horowitz, was unambiguous when the round was announced: “Financial institutions are sitting on decades of infrastructure debt, and that debt is now the single biggest obstacle to AI adoption. What Stitch is building, a modern, unified system of record, is what makes everything else possible. We’re excited to support them, and honoured to make this our first investment in the region.”
The firm went further in its investment blog, describing the Middle East as a rare greenfield for exactly this kind of company, where “a nationwide focus on expanding outside of oil is feeding one of the fastest-growing financial services ecosystems in the world,” with the industry growing at 10% year on year. The GCC fintech market, valued at over $10bn in 2025, is projected to reach nearly $30bn by 2032. Saudi Arabia plans to establish hundreds of new financial institutions in the coming years, each of which will need modern, scalable core infrastructure from the start. After years of building conviction around the global fintech infrastructure opportunity, a16z described Stitch as “the clearest expression we’ve found of what the next generation of this infrastructure looks like.”
The $25m Series A, joined again by Arbor Ventures, COTU Ventures, Raed Ventures, and SVC, all of whom backed the $10m seed round, will fund product development, deeper market coverage across the GCC and wider MENA, and the early stages of a global go-to-market operation. Stitch currently operates in Saudi Arabia, the UAE, the Philippines, Egypt, and Kenya.
The Value That Comes After the Foundation
The commercial traction is visible in the numbers: more than $5bn transacted on the platform in the six months to May 2026, tenfold customer growth across 2025, and twentyfold revenue growth in the same period. Most enterprise engagements run to six figures at minimum and frequently into seven. For most customers, Stitch is also their first cloud deployment. The infrastructure is mission-critical in the most literal sense: if Stitch stops, the customer stops.
The AI opportunity atop that foundation is what the current investment is really about. As a16z noted, Stitch’s platform is positioned to support financial institutions in their AI transformation by adding AI-powered customer support, automating loan origination, and eliminating the manual inbox workflows that consume operational headcount at every institution. Oueida described a demo held days before the funding announcement in which an existing client, watching an AI feature run on their own data within the Stitch platform, asked to go live immediately. Not a demonstration on synthetic data, not a chatbot fielding isolated queries, but genuine operational efficiency running across workflows that matter. “This is true value extraction,” Oueida said.
The global ambition is stated without qualification. The same fragmentation problem in Riyadh exists in São Paulo and Jakarta as well. Stitch plans to enter each new market the way it has approached the ones it already operates in: deep partnerships, direct engagement with regulators, and no shortcuts on compliance.
