For the moment you're
deciding whether to email me back.

A banking product 10,000 customers already asked for.

Lily is the financial operating system for the entrepreneurs the mainstream banking system refuses to serve — built on stablecoin rails, wrapped in a multilingual AI, distributed through a profitable six-year-old business that already serves them.

I've been running Launchese — a UK company-formation business for international founders — since 2019. Profitable every year. No outside capital. Ten thousand paying customers. For six of those years, our support inbox has been a counselling service for people whose Stripe, PayPal, Mercury, Wise, or Revolut account just got closed.

I built Lily solo over six months because I finally got tired of telling people I couldn't help. Banking is shipped. AI assistant is shipped. Voice mode is shipped. 214 tests green. Now I need a co-founder to take this from "one person who built the thing" to "a company that wins the category."

This is the mode, not the edge case.

From six years of support tickets, Reddit deep-dives, and weekly 1:1 conversations with our customers in Turkey, Egypt, Nigeria, the UAE, and a hundred other countries:

These aren't outliers. 78% of SMEs still depend on traditional banks for cross-border payments. 55% complain about unfair pricing. 65% of entrepreneurs in a recent UAE study said banking was the biggest challenge during setup. Stablecoin transaction volume hit $33 trillion in 2025. The international entrepreneur has been quietly begging for a fix for ten years.

The first quote is the one I want you to remember. Most people think this is a "high fees" problem. It isn't. For 60–70% of entrepreneurs in emerging markets, it's a pre-revenue blocker — they cannot start earning at all.

It's structural, not competitive.

Every neobank you've heard of is built for a different customer. Their risk models were trained on Western fraud patterns and treat any non-Western profile as elevated risk. Hostile by design, not by accident.

None of them is going to wake up tomorrow and decide to serve our customer. The people they freeze are not a bug in their model. They're the model working as designed.

Six years of paying customers no one else has access to.

For six years Launchese has been operating, by accident, as the concierge tier of a bank we do not own. People pay us to form their UK Ltd, handle the registered address, manage the company secretary — and then call us in tears when Stripe terminates them.

Lily is the banking product I should have built four years ago. The acquisition cost for the first 5,000 banking customers is an email from me.

The problem isn't banking. It's trust.

Our customers already hold money in USDT. They already open accounts at three providers as insurance. They are sophisticated. They are not asking for technology. They are asking for a bank that picks up the phone when something goes wrong.

Here's the part most fintech founders miss. The entrepreneurs who've been rejected by Wise, Stripe, and PayPal are not price-sensitive. They're gratitude-sensitive. You solve their problem; they stay for life. That changes the unit economics, the churn assumptions, and the competitive posture. It's why we won't have to compete on fees with Cenoa or Slash. We compete on serving the underserved — and we win the moment we are the first ones who didn't kick them out.

Launchese is already that relationship. Lily extends it to banking. The product is built on Bridge (Stripe's stablecoin infrastructure, which they bought for $1.1B in 2024) — but the feature that matters is the human at the other end of the line. Not an email. Not a ticket. A phone call from someone whose name they already know.

Meet Sentinel.

Every other bank uses AI to police the user — to find reasons to ban them. Lily uses AI to protect the user — to find ways to keep them safe. That single inversion is the most defensible thing we're building.

Underneath this sits a Document Factory. Sentinel auto-generates the bank-grade paperwork most international entrepreneurs don't have but compliance officers expect — Operating Agreements, Share Certificates, AML manuals, Source of Wealth declarations, Master Service Agreements, Board Resolutions, Inter-Company Loan Agreements, Click-Wrap evidence packs for SaaS subscription revenue.

This is the thing none of the well-funded competitors have. Slash is a USD account. Dakota is an API. Cenoa is virtual accounts in emerging markets. Meow is high-yield checking for US tech startups. Every one of them treats compliance as a tax their users pay. We treat it as a premium product. "We built an AI to keep you safe" is a different category than "we built a faster way to send money."

And it's a real moat. Three years from now, Sentinel will have processed millions of compliance events across thousands of jurisdictions. That data is the rule library. Slash and Dakota have to start that clock today; we already have a head start because Launchese has been doing this manually, by hand, for six years.

This is not a deck-only company.

Six months. One person. Built in public. Here's what's real and what isn't.

Big enough, but that's not the question.

Cross-border B2B flows are $31.7 trillion a year. SMEs are 43% of that — $13.8 trillion. The fintech-addressable slice is $4.8–6.9T in flow, generating $62–90B in annual revenue. The narrower stablecoin-ready segment is $3.1–9B in revenue. We call our SAM $8.9B, our SOM at three years $44–89M. Pick whichever number you trust.

The real question is not "is the market big enough." It's "can we execute fast enough." Year one — converting only existing Launchese customers, no paid acquisition — is $1.3M–$2.3M ARR. That's the floor, not the ceiling. The corridor map underneath is even better: Nigeria→UK alone is a $40B annual diaspora flow at 8% average cost. Cut that to 2% and we save Nigerian families collectively a quarter of a billion dollars a year.

Five revenue lines. All margin-positive on day one.

Lily isn't a single-take-rate business. The product is layered, so the unit economics are layered. None of these depend on a future hand-wave — they're priced against Bridge's published costs today.

On top of that sits the platform plan ($29–$49/month) and the existing Launchese formation revenue ($149–$499 one-time). The blended ARPU comes out at $25 in Year 1 climbing to $45 by Year 5, with gross margin north of 70% from Year 2 once Bridge volume discounts kick in. Year 3+ optionality — supply chain finance, business credit underwritten by on-chain transaction data — opens a $5T B2B factoring market we don't need to count today.

Real category. Real competitors. An honest map.

Eight well-funded stablecoin neobanks are now in this space. The deck I sent you names Slash and Dakota. They're not the closest threat. Cenoa is.

Cenoa is the head-on play. Same Bridge infrastructure. Same emerging-market focus. They scaled to 10,000 users and $5M monthly volume in six months. Their seed round was led by Quiet Capital, Human Capital, Ulu Ventures, Acrew Capital. They are not theoretical.

So why do we win? Because we start at 10,000 customers, not zero. Cenoa is acquiring strangers; we are converting people who already pay Launchese for company formation, who already trust us with the most sensitive part of their setup. We add UK formation as a wedge they don't have. We add 14 languages at launch where they have one. We add Sentinel where they have a virtual account. Same rails, fundamentally different product.

Slash and Dakota are also real, but they're not coming for our customer. Slash is a general business tool. Dakota is a developer-facing API for institutions. Neither speaks Turkish, neither has formation services, neither has a founder who has spent hundreds of hours on the phone with these specific entrepreneurs.

Four forces converging. The window closes by 2028.

The honest version.

Every co-founder I'd want to work with is going to ask. Here's my answer in advance.

Three profiles. In order of where the gap is biggest today.

Not looking for: another full-stack engineer, a consultant or advisor, or a pure fundraiser. The work is operational, regulatory, or design — in that order. We start with a paid three-month pilot (see Section 14), so you don't have to leap before you've looked.

In two stages, not one leap.

Most co-founder pitches ask for a four-year leap of faith on day one. I think that scares away the people I most want to talk to — the ones who already run things. So we do it differently.

The first three months are a paid pilot. You don't quit anything you're already doing. We both find out if this is real before either of us makes a longer commitment. There's a meaningful financial commitment from my side during the pilot — I'd rather walk you through the specifics on our first call than negotiate them in a document, but it is real and it is on the table.

Phase 1 — The pilot (Months 1–3)

Phase 2 — Co-founder (if both still want it)

The reason this is two stages is simple. A bad co-founder fit costs both of us about a year of life. A paid three-month pilot is a much cheaper way to find out — and a much more generous way to start a serious conversation.