The infrastructure and code in financial services is like sedimentary rock. Each layer builds on top of the other. You can’t have a deposit product without a bank charter. You can’t have a card without Visa or Mastercard. You can’t move money internationally without SWIFT.

This is the fintech stack.

Five layers. Each one dependent on the one below.

Understanding these layers is like getting X-ray vision into how finance actually works. Most people see the app. They don’t see the geology beneath it.

Brought to you by Fintech Brainfood: The weekly newsletter on all things Fintech. Hit subscribe to stay up to date.

At the bottom, you have rails. The pipes that move money.

These are the networks that actually transport value from one place to another. Think of them as the financial internet – except most of it was built before the actual internet existed.

Card networks: Visa and Mastercard process the overwhelming majority of card transactions globally. When you tap your phone at Starbucks, your payment travels through their network. AmEx and Discover run their own closed-loop networks. In China, UnionPay dominates.

Bank-to-bank systems: ACH (Automated Clearing House) in the US handles direct deposits and bill payments. In the UK, it’s Faster Payments. SEPA for Europe. Each country has its own domestic clearing system, often running on infrastructure built decades ago.

International wires: When you send an international wire, SWIFT carries the instruction. The actual money moves through correspondent banking relationships. It’s slow. It’s expensive. It’s remarkably resilient. But SWIFT itself doesn’t move money. Banks do.

Real-time payments: FedNow in the US launched in 2023 after years of delays. The UK has had real-time payments since 2008. India’s UPI processes billions of transactions monthly. This is where the action is – instant, 24/7 money movement.

The new rails: Blockchain networks like Ethereum and Solana. Stablecoins like USDC and USDT. These are genuinely new financial infrastructure. I believe Stripe’s acquisition of Bridge will go down as an all-time great fintech acquisition because it recognized this shift. Stablecoins are the first net-new payment rail in decades. They’re 10x more efficient for cross-border settlement. That’s the repricing.

Above the rails, you need capabilities. The operational machinery that makes financial products compliant and safe.

This is where most people’s eyes glaze over. But here’s the thing – this layer determines what products are even legal to offer.

Identity verification (KYC): Know Your Customer. Before you can open an account, someone has to verify you are who you claim to be. Companies like Onfido, Jumio, Socure, and Persona do this. Most consumers have no idea this step exists. They just know uploading their driver’s license was annoying.

Anti-money laundering (AML): Detecting suspicious transactions. Filing SARs (Suspicious Activity Reports) with regulators. This is where Sardine, ComplyAdvantage, and Chainalysis live. Banks spend billions on this annually. The dirty secret? A 2022 report showed less than 1% of illicit funds are actually caught by AML systems. We need to do better.

Fraud prevention: Stopping bad actors from stealing money or creating fake accounts. Companies like Sardine, Sift, and Forter analyze signals – device fingerprinting, behavioral biometrics, transaction patterns – to catch fraud before it happens.

Compliance and licensing: Understanding what regulations apply and maintaining the licenses required to operate. This is increasingly powered by AI. Regtech companies like Comply and Alloy help automate what used to require armies of compliance officers.

Credit decisioning: Determining whether someone should receive a loan and at what rate. This is where FICO scores originated. Now companies like Plaid, Experian, and Nova Credit pull alternative data – bank transactions, rent payments, employment verification – to build more complete pictures.

The capabilities layer is increasingly AI-native. Large language models are remarkably good at parsing regulatory text, analyzing transaction patterns, and automating compliance workflows. Legacy finance is perfect for AI because nearly all money moves as structured text files.

Now we get to the building blocks consumers recognize.

Products are the fundamental financial instruments. Not the app. Not the brand. The underlying financial mechanics.

Deposits: Checking and savings accounts. Money market funds. CDs. Anywhere you can park cash and (hopefully) earn interest. JPMorgan Chase, Bank of America, and Wells Fargo are the giants. But increasingly, fintech companies offer deposits through partner banks.

Lending: Mortgages. Auto loans. Personal loans. Credit cards. Lines of credit. This is where money gets created (fractional reserve banking – banks lend out more than they hold in deposits). LendingClub, SoFi, and Upstart have built lending-first businesses. Block’s Cash App and PayPal do embedded lending.

Cards: Debit cards (spend money from your account). Credit cards (borrow money with each swipe). Prepaid cards (spend pre-loaded funds). Companies like Marqeta, Lithic, and Highnote provide the infrastructure for others to issue cards.

Payments: The ability to send, receive, and process money. Stripe and Adyen for merchants. Plaid and MX for connecting to bank accounts. Square (now Block) for point-of-sale. PayPal for peer-to-peer.

Investments: Brokerage accounts. Retirement accounts. Robo-advisors. Crypto trading. Robinhood democratized stock trading. Wealthfront and Betterment automated portfolio management. Coinbase opened crypto to the masses.

Insurance: Protection products. Life. Health. Property. Auto. Lemonade brought AI underwriting to renters insurance. Root uses telematics for auto insurance. Embedded insurance is sneaking into everything.

Products are Lego bricks. The same deposit product can be packaged completely differently depending on the brand and use case.

Services are where products get packaged for specific outcomes.

This is where fintech companies differentiate. Same underlying products, radically different positioning and features.

“Get paid early”: Chime, Dave, and Earnin all offer early wage access. The underlying product is just a direct deposit prediction. But the service – getting your paycheck two days early – feels magical to someone living paycheck to paycheck. Poor communities need a hand up, not a handout. A great product is a hand-up.

Buy Now, Pay Later (BNPL): Affirm, Klarna, and Afterpay turned point-of-sale lending into a consumer-facing service. The underlying product is an installment loan. The service is “split this purchase into four payments.” I’ll die on this hill: BNPL is good. It’s transparent about costs and helps consumers budget.

Expense management: Brex, Ramp, and Airbase took corporate cards and added spend management software. The underlying product is still a credit card. The service includes receipt capture, approval workflows, and accounting integrations.

Treasury management: Mercury, Rho, and Novo offer startups business banking with automated sweep networks. The underlying product is deposit accounts. The service is higher yields and better UX than traditional banks.

Vertical-specific banking: Relay for small businesses. Found for freelancers. Greenlight for kids. Same banking products, purpose-built for specific audiences.

The service layer is where marketing meets product. It’s where “we offer checking accounts” becomes “we help freelancers manage irregular income.”

At the top of the stack, you have journeys.

Not products. Not services. Life moments.

This is where financial services meet the real world.

Buying a car: You need lending (auto loan), insurance (auto policy), and payments (monthly installments). Today, you deal with three different companies. Tesla integrated all of it. That’s the vision of embedded finance.

Starting a business: You need incorporation, banking, payroll, accounting, and possibly lending. Companies like Stripe Atlas and Doola collapse these into unified experiences. Mercury and Brex compete on being the primary financial relationship.

Buying a home: Mortgage pre-approval. Title insurance. Homeowners insurance. Property taxes. Moving costs. Better.com and Blend tried to simplify this nightmare. It’s still a nightmare. Massive opportunity.

Paying rent: Security deposit. Monthly payments. Renters insurance. Credit building. Startups like Bilt turn rent payments into rewards points and credit history.

Getting divorced: Division of assets. New bank accounts. Updated beneficiaries. Financial planning. Almost no fintech addresses this directly. Wild, given how common it is.

Managing aging parents: Power of attorney. Joint accounts. Fraud monitoring. Estate planning. Carefull and others are barely scratching the surface.

The journeys layer is where the biggest opportunities remain. Most fintech is still solving individual product problems, not whole life moments. The companies that crack this – that truly own a journey end-to-end – will be the trillion-dollar fintech companies of the next decade.

Understanding the fintech stack gives you a map.

When you see a new fintech startup, ask: which layer are they playing at?

The layers are also where disruption happens. AI is transforming the capabilities layer right now. Stablecoins are creating new rails. Embedded finance is collapsing the service and journey layers into the places where life happens.