Drop #5 now live, and it's rebuilding financial infrastructure for European startups

Hello friends, and welcome to The Unsophisticated Investor! Brought to you by Scott & Rob, the founders of Shuttle!

We’re back with the first investment Drop of 2026! It’s a company we've been watching closely for a while. A team of former operators from some of the most consequential fintechs built in the last decade, backed by one of Europe's leading B2B venture funds, building what they believe is the financial home for European startups. The product is live. Early retention numbers are the kind that make investors pay attention. And the core thesis, that the financial infrastructure serving Europe's most dynamic companies is broken in a way that nobody has properly fixed, is one we find genuinely compelling and can relate to.

If you want to invest alongside the VC funds who've backed breakout companies like Revolut, Asana, JustEat, Bolt, Lets Get Checked, Loom, Runna, Charlotte Tilbury, Deel, Aircall, AngelList, Carta, TransferWise and many more, regardless of your knowledge, network or net worth, you can signup today by clicking the button below 👇

Now, let’s get into it 👇

The story everyone tells about AI and fintech

Ask someone working in financial services what AI is doing to their industry and you'll get a version of the same answer. Fraud detection is sharper. Credit models are more accurate. Customer service bots are handling queries that used to require a human. Back-office processing is faster. Compliance monitoring is more automated.

All of this is true, but it’s not the interesting part.

What's being described is AI applied to financial services as they already exist: the same processes, the same workflows, the same underlying logic, running on better models. Fraud detection was always about pattern recognition. AI does pattern recognition better than humans. Credit scoring was always about predicting default probability. AI does that better too. The work is the same. The tool is faster.

This is a meaningful improvement. It is not a transformation. And the distinction matters enormously for where the real value in fintech is about to be created.

What transformation actually looks like

There is a different category of AI application that barely features in the financial services conversation, because it doesn't look like a feature. It looks like the absence of a job.

Consider what it means to run the finances of a VC-backed startup in Europe today. You have a bank account at an incumbent that processes things in batch cycles and takes three weeks to open. A neobank account for speed. A payroll provider. Something for expenses. Something for FX. Possibly a separate tool for invoicing. A bookkeeper, in-house or outsourced, who reconciles all of it manually and delivers financial visibility two to four weeks after month-end. The average VC-backed founder, building a company that may well be developing AI products itself, is still processing invoices by hand and waiting a fortnight to know whether they can make payroll.

This is not a technology problem in the sense of lacking the right software. Every one of those tools has an API. The problem is that nobody has rebuilt the workflow itself around what's now possible.

An AI system with access to a founder's inbox, banking connections, and accounting integrations doesn't need to speed up invoice processing. It can eliminate it as a human activity. The invoice arrives in email, gets extracted automatically, matched to a transaction, and queued for one-click approval. Payroll doesn't need a faster interface for entering figures. It can read the accountant's PDF and generate the bulk payment file. Month-end reconciliation across four bank accounts doesn't need better spreadsheet tooling. It stops being a month-end event. The management accounts exist continuously, updated in real time, investor-ready without anyone producing them.

This is the difference between AI as acceleration and AI as replacement. The first makes existing processes cheaper. The second makes certain processes unnecessary.

Why the neobanks couldn't get here

The obvious question is why this hasn't already happened. The neobank wave was supposed to fix business banking. Revolut, Monzo, Qonto, Tide, a decade of well-funded challengers, and the average VC-backed founder is still stitching together five tools and waiting a fortnight for their numbers.

The answer is structural, not technical.

Neobanks are built for volume. The economics work when you have millions of customers, most of them simple, paying relatively little per month. That model is fundamentally incompatible with the depth of service that a complex, fast-growing startup actually needs. Dedicated relationship managers who understand multi-entity corporate structures. Multi-currency treasury that doesn't require a workaround. Sophisticated investor reporting that a founder can hand to their board without apology. These capabilities don't survive contact with a business model optimised for micro-business scale.

Point solutions went in the opposite direction and created a different version of the same problem. Brex went deep on corporate cards and expense management. Pleo on expenses. Ramp on spend control. Each solved one workflow and left the founder with another contract, another login, another data silo that doesn't talk to the others. The fragmentation didn't shrink. It just became more sophisticated.

The companies that could have solved it, the traditional banks, are running on technology that predates the iPhone and serving a customer base that doesn't demand better. They sit in almost every startup's banking stack, doing nothing more than holding cash that can't be moved anywhere quickly enough to be useful.

Why this time the architecture is different

The window for fixing this has existed theoretically for years. What's changed is that AI has moved the build cost.

Rebuilding the financial workflow for a complex startup from scratch used to mean either hiring an army of operations staff or convincing founders to tolerate a product that covered sixty percent of their needs. AI collapses the operations requirement. A platform that would have required dedicated human capacity to handle invoice matching, payroll processing, and account reconciliation at any meaningful scale can now do those things with minimal marginal cost per customer. That changes the unit economics of serving a high-complexity customer segment in a way that simply wasn't available to the previous generation of challengers.

But, and this is the part that gets missed in most AI-in-fintech conversations, the architecture has to be built for this from the ground up. Legacy rails don't benefit from being connected to a better model. The workflows have to be redesigned around what AI can actually do, not retrofitted around what humans used to do. Which is why the companies that matter in this category are the ones building now, not the ones trying to adapt what they built in 2017.

The regulatory layer nobody talks about

There is a second part of the architecture that matters as much as the AI layer and gets considerably less attention: the regulatory stack - something we are very familiar with here at Shuttle. 

Building a genuine financial platform in the UK and European market requires, at minimum, regulatory authorisations, Visa or Mastercard approval for card issuance, and a credible path towards licensing for the accounts and payments layer. Each of these takes time, operational credibility, and a working relationship with regulators that cannot be shortcut by a better model or a faster shipping cycle. The compliance infrastructure is not overhead. It is a competitive moat, because it sets a floor on how quickly a technically superior competitor can enter the same space.

We wrote last week about how regulatory moats are structural defensibility in the AI era. This is the clearest example in European fintech. The companies that do the regulatory groundwork at pre-seed and seed stage are building an asset that compounds in value as they scale, and that is functionally impossible for a new entrant to replicate without starting the clock again.

Our take on what this means for private markets

The consumer neobank story is largely told. The winners are known, the valuations are eye-watering, and the debate about whether they justify themselves will run for years.

The B2B financial infrastructure story for startups is earlier and structurally different. Higher switching costs: once your payroll, treasury, invoicing, and accounts are on one platform, leaving means rebuilding four workflows simultaneously. Predictable, diversified revenue that isn't entirely dependent on interest rates or interchange economics. A customer base that grows in value as the companies it serves grow. And unit economics that improve materially as the platform replaces third-party banking infrastructure with its own licences.

The private market principle here is straightforward. Categories get defined in private markets, at seed and early growth stage, before the winner is obvious. By the time the thesis is confirmed and the round is oversubscribed, you are not investing in the opportunity. You are paying for other people's conviction. The investors who matter in the B2B startup financial infrastructure category are making that call now.

The team building this week's deal understands the architecture, they've built parts of it before, at companies that are now the infrastructure everyone else builds on top of. That's a different starting point than another neobank with a cleaner app.

The bottom line

AI is transforming financial services. But the transformation worth paying attention to isn't faster fraud models or smarter credit decisions. It's the elimination of entire categories of financial work that have been consuming founder time and operational budget since the first startup opened a bank account.

The companies getting this right are not building better versions of what exists. They are building on a different premise entirely: that the financial operating system for a modern startup should require approximately zero manual input to produce complete, real-time visibility of the business. That's a bigger claim than it sounds. And the window to back it at seed stage is, by definition, short.

What we’ve been working on at Shuttle

• Working on the pipeline for our next drop 🚀

• Deep in fundraising mode as we kick-off our pre-seed raise 👀

• Shipping some really exciting AI product features ✨

The Unsophisticated Investor is brought to you by Scott & Rob, the founders of Shuttle. We’re both sick of private markets being a playground exclusive to the ultra-wealthy so we started a company to challenge the status-quo. Shuttle’s singular focus is to unlock private markets for Millennial and Gen Z tech professionals and help them build wealth through the highest performing private market opportunities.

Scott & Rob
Shuttle Co-Founders