Payments infrastructure · Founded 2010 · Private · $159B as of the February 2026 tender
Stripe
The payments company developers reach for first, with online card acceptance in a few lines of code.
The problem they saw
In 2010, accepting a credit card online meant a merchant account, a payment-gateway
contract, and a multi-week integration across a bank, a processor, and a fraud system
before a single dollar could move. Brothers Patrick Collison (born 1988) and John Collison
(born 1990), in their late teens and early twenties at the time, kept watching developer
friends abandon side projects at exactly that step.
The founding insight
For Patrick and John, payments was not a banking problem to pitch to a CFO. It was a
developer-experience problem to solve for the person writing the code. Make it as easy as
pasting in a snippet, and developers would bring the volume to you instead of you chasing
merchant accounts one contract at a time.
What version 1 actually was
Seven lines of code. A copy-paste integration snippet that let a developer accept a card
payment in about a day. Patrick and John personally handled the bank and processor
relationships behind the scenes, so nobody using it ever had to see that mess.
How it makes money
Stripe takes a cut of every successful transaction, at a standard published rate of 2.9%
plus 30 cents per charge for US cards, and stacks priced add-ons like billing, fraud
protection, and treasury on top. Revenue scales with both volume and how much of the stack
a business adopts.
Where it is now
Stripe processed $1.9 trillion in total payment volume in 2025, up 34% year over year,
and says it is robustly profitable. An employee tender offer announced in February 2026
valued the company at $159B, with no imminent plans to go public as of that date.
The student-sized version
- Problemsomething painful takes too many steps for the person actually trying to build with it.
- Userthe developer, or the student, doing the work, not the executive who signs the contract.
- V1a stripped-down, unglamorous piece of glue code that removes just the worst step, with everything else handled by hand behind the curtain.
That is a Rocky Top entry: one problem, one user, one small buildable thing.
Consumer credit, student credit-building · Founded 2021 · Private · $14.4M seed in 2024
Mine founded as Fizz
A credit-building debit card built by two college dropouts for students who cannot get a real credit card yet.
The problem they saw
Carlo Köbe and Scott Smith saw the wall every college student hits: you need credit
history to get approved for a credit card, and you cannot build credit history without
one. Most students get rejected, get stuck with a parent as cosigner, or put the whole
thing off until after graduation, which leaves them years behind on something that governs
rent applications, phone plans, and loan rates for the rest of their lives.
The founding insight
A credit-building product for students does not have to look or work like a real credit
card, and it should not carry the late fees and interest charges that make one risky for
somebody who has never managed a balance before. Give students a debit-card experience
that reports to the bureaus the way an issuer would, and the credit-building benefit
arrives without the part that gets people in trouble.
What version 1 actually was
Köbe and Smith left Harvard and Cornell, at 19 and 21, and went through Y
Combinator's Summer 2021 batch. The artifact was a debit card rather than a credit card:
no interest, no late fees, nothing hidden, built to report to the bureaus so students
could start a credit file using a card that could not put them into debt.
How it makes money
Revenue comes from interchange on card transactions, affiliate and brand-partner deals,
optional subscription products, and partnerships with universities and merchants. The
model is built around usage rather than around penalizing a customer for missing a
payment.
Where it is now
The company raised a $14.4M seed round led by Kleiner Perkins, with Y Combinator, SV
Angel and others participating, announced June 2024. It reported more than 150,000
registered users at that point, available at over 300 colleges and universities. It has
since rebranded to Mine and widened past credit building into a fuller money app with
budgeting and insights. No funding round newer than that seed had surfaced as of August
2026.
The student-sized version
- Problema financial product locks you out precisely because you are new to finance, the classic chicken-and-egg gap.
- Usersomebody at exactly the founders' age when they started, with no credit history and no straightforward way to get one.
- V1strip a real product down to the single feature that matters, which is bureau reporting, and remove the part that makes it dangerous, which is debt.
That is a Rocky Top entry: students building for students, on a problem they were personally stuck inside.
Digital banking · Founded 2013 · Public, NYSE: NU · roughly $70–71B as of August 2026
Nubank
The purple credit card that took on Brazil's five-bank oligopoly, with no branches and no lines.
The problem they saw
David Vélez tried to open a Brazilian bank account and spent roughly 45 days on
paperwork. He was not alone in that. Five institutions controlled most of the country's
retail banking, credit-card interest rates routinely topped 400% APR, and almost no
low-cost alternative existed.
The founding insight
Vélez, Cristina Junqueira and Edward Wible read the situation as bigger than a
slightly-better-bank fix. Brazil's banking market was a slow oligopoly with no reason to
compete on speed or price. So they built an institution with zero physical branches and
made being entirely app-based the whole pitch.
What version 1 actually was
A translucent purple, no-annual-fee Mastercard managed completely through an app, with
nowhere to walk in. Instead of buying ads, the company grew through an invite-only velvet
rope: existing users invited friends, and a waitlist passed one million people in about 18
months before most of them had ever held the card.
How it makes money
Interchange on card transactions, paid by merchants rather than customers, got the
business started. Interest income from revolving balances and personal loans has since
taken over as the lending book scaled, reaching about 83% of revenue by 2024, up sharply
from 2020 when fees and commissions made up roughly half.
Where it is now
More than 135 million customers globally as of Q1 2026, over 115 million of them in
Brazil, adding roughly 4 million in that quarter alone. Market capitalization sits around
$70–71 billion in USD as of August 2026, per Nu Holdings' SEC filings, with trackers
landing within a point or two of each other.
The student-sized version
- Probleman entire system is slow and expensive because nobody inside it has to compete.
- Usereveryone stuck in that system, whether or not the incumbents already serve them well.
- V1strip the product down to the one thing that breaks the pattern, then grow it by making current users the whole marketing plan.
That is a Rocky Top entry: find where the incumbents have no reason to move fast, and be the thing that moves fast.
Personal finance, subscription management · Founded 2015 · Subsidiary of Rocket Companies, NYSE: RKT · acquired December 2021 for $1.275B
Rocket Money founded as Truebill
Three brothers built a tool to cancel their own forgotten subscriptions, then found everyone else had the same junk drawer.
The problem they saw
In 2015, brothers Haroon, Yahya and Idris Mokhtarzada kept spotting charges on their own
bank statements for things they had forgotten they were paying for. Free trials that had
quietly turned into subscriptions, memberships nobody remembered joining, small recurring
fees each too minor to chase and, added together, not minor at all.
The founding insight
They zeroed in on cancellation as the part worth solving. Nobody needed help seeing that
they were bleeding money in small increments. What people needed was help actually getting
rid of the charges, so the brothers built the tool that does the canceling instead of one
more dashboard showing you the damage.
What version 1 actually was
Started in Silver Spring, Maryland, and put through Y Combinator's Winter 2016 batch, the
company began as something the brothers made to fix their own problem before any company
existed around it: a way to surface recurring charges on a linked bank account and cancel
whichever ones you did not want.
How it makes money
Two revenue lines. A pay-what-you-think-is-fair Premium subscription, roughly $7–$14
a month, unlocks cancellation, full credit reports and custom budget categories. Separately,
when the company negotiates down a user's bill it keeps 35% to 60% of the first year's
savings as a success fee, so that half of the business only earns when it saves you
something.
Where it is now
Rocket Companies acquired Truebill for $1.275B in cash, announced December 20, 2021, when
it had 2.5 million members, and renamed it Rocket Money in August 2022. The product reports
more than 10 million members as of August 2026, over $2.5B saved cumulatively, and a
separate milestone of one million paid Premium members.
The student-sized version
- Problemsmall recurring costs that are individually forgettable and collectively real, easy to miss until somebody points them out.
- Userliterally the founders themselves, before it was anyone else's problem too.
- V1a tool built to fix your own annoyance, with no company attached to it yet.
That is a Rocky Top entry: build the thing that solves your own problem, then find out whether it is everybody's.
Personal finance, comparison marketplace · Founded 2009 · Public, NASDAQ: NRDS · roughly $636M as of August 2026
NerdWallet
A spreadsheet built for one person's sister, turned into the site people check before picking a credit card.
The problem they saw
In 2009, Tim Chen had just been laid off from a hedge-fund analyst job when his sister
emailed him something simple: which credit card should she get if she wanted to avoid
foreign transaction fees while traveling? Chen worked in finance for a living and it still
took him a week of reading fine print to answer her properly. If it took him that long, it
was taking everybody else longer.
The founding insight
The hard part of picking a financial product was that the comparison itself lived
nowhere. Issuers, banks and lenders each published their own terms in their own format,
and no one had built the single place to line them up side by side. Build that place, and
the comparison becomes the product.
What version 1 actually was
An Excel spreadsheet, made to answer his sister's question. Chen kept adding to it as
more people asked him similar things, and it began circulating past his own circle. He
seeded the resulting company with $800 and worked unpaid through the first year while
turning that file into a website.
How it makes money
Readers are not charged. Commissions and fees come from the banks, issuers, lenders and
insurers being compared, paid when a user clicks through, applies, gets approved or funds
an account. A NerdWallet+ subscription adds a second, consumer-paid line on top. Worth
noticing that those two models pull in different directions, since one is funded by the
reader and the other by the product under review.
Where it is now
The company went public on Nasdaq in November 2021 at a $1.17B valuation. Market
capitalization sits at roughly $636M as of August 2026, down about 45% from that listing.
It reports 15 million registered members as of 2024, which is the figure worth citing,
because third-party traffic estimates measure visits rather than members and run
lower.
The student-sized version
- Problema decision that should take ten minutes takes a week, because what you need in order to compare options is scattered and dense.
- Usersomebody who has to make the choice without being a finance professional, so a sibling, a classmate, or you.
- V1a spreadsheet rather than an app, just the comparison, built for one real person's real question and useful to far more people than the one it was for.
That is a Rocky Top entry: solve it for a single person first, in the simplest tool you already know.
Banking data infrastructure · Founded 2013 · Private · $8B as of the February 2026 round
Plaid
The plumbing behind the "connect your bank account" button on most fintech apps.
The problem they saw
Zach Perret and William Hockey spent 2013 trying to build consumer budgeting and
bookkeeping tools, the kind that has to see your balance to be worth anything. Every
single time, the hardest part was getting bank data to show up at all.
The founding insight
Perret and Hockey came to see their product differently. Not a budgeting app, but the
missing plumbing sitting underneath every budgeting app. Rather than competing to make the
best consumer tool on top of broken connections, they built the connection itself and sold
it to every other developer hitting the same wall.
What version 1 actually was
A scrappy bank-activity mapping demo called Rambler, made to prove the connection worked.
Rambler won the 2013 TechCrunch Disrupt hackathon in Manhattan, and that win was the pivot
point that sent both founders to San Francisco to build an API business instead of another
consumer app.
How it makes money
Pricing runs per successful API call, roughly 10 to 60 cents depending on product and
volume tier, with account-verification calls running 30 cents to a dollar per successful
connection, plus monthly or annual minimums on enterprise contracts. Revenue tracks how
much bank data other companies' apps actually pull.
Where it is now
Plaid connects to more than 12,000 financial institutions across the US, Canada, the UK
and Europe, with over 9,000 apps built on top. The company says one in two US adults with
a bank account has used something connected through it. A funding round valued it at $8B
as of February 2026, still below its 2021 peak of $13.4B.
The student-sized version
- Problemtwo different builders keep hitting the exact same wall before they can even start on their real idea.
- Userthe next developer who will build on top of you, since the end consumer is not who this is for.
- V1a scrappy demo, a hackathon build in this case, proving the underlying connection works before any real product surrounds it.
That is a Rocky Top entry: find the wall everybody hits, then build the thing that gets them past it.
Consumer banking, neobank · Founded 2012 · Public, NASDAQ: CHYM · roughly $11.9–12.1B as of August 2026
Chime
The no-fee account built for people who kept getting hit by overdraft charges at the banks they already had.
The problem they saw
Chris Britt, who had worked at Visa and Green Dot, and Ryan King watched one story repeat
endlessly. People living paycheck to paycheck were taking overdraft fees and
minimum-balance charges from institutions supposed to help them manage money, and instead
penalizing them for having too little of it.
The founding insight
An account does not have to earn its keep by charging the customers least able to pay. If
the business made money the way a retailer does, taking a small cut of what people spend
rather than a penalty for falling short, its incentives would point the same direction as
the customer's instead of against them.
What version 1 actually was
Founded in 2012, the product did not publicly launch until April 2014, including an
appearance on the Dr. Phil show to get the word out. A no-fee checking account and a debit
card were the entire offering, built to prove a bank could survive without overdraft
revenue at all.
How it makes money
No monthly fees, no overdraft fees, no minimum balance. Roughly 80% to 90% of revenue
comes from interchange instead, a cut of about 1.5% that merchants pay on every debit
swipe, so the company only earns when members spend on the card.
Where it is now
10.2 million active members as of Q1 2026, up 19% year over year and up from 9.5 million
at the end of 2025. The company listed on Nasdaq as CHYM in June 2025, and its market
capitalization clusters around $11.9–12.1 billion across trackers as of August
2026.
The student-sized version
- Probleman entire industry's revenue model runs on charging its most vulnerable users.
- Userthe person the existing product profits from rather than serves.
- V1strip out every fee and find out whether a simpler, aligned incentive can still pay for itself.
That is a Rocky Top entry: find a business model built against the people using it, then build the version that is not.
Corporate cards, business spend management · Founded 2017 · Acquired by Capital One · $5.15B, closed April 2026
Brex
Two 22-year-olds pivoted out of a VR startup because they could not get a corporate credit card of their own.
The problem they saw
Henrique Dubugras and Pedro Franceschi, Brazilian entrepreneurs both around 22, entered Y
Combinator's Winter 2017 batch with a virtual-reality idea. Three weeks in they hit an
unrelated wall: no bank would give their own company a corporate card. Every one wanted a
personal guarantee and a credit history neither founder had, which is the same
young-person problem wearing a business suit instead of a student ID.
The founding insight
Banks were underwriting startup cards against the wrong thing. A pre-revenue company with
venture funding in the bank has real money and a real ability to pay. What it lacks is the
personal credit history or years-in-business a traditional issuer wants to see. Underwrite
against cash in the account rather than the founder's credit score, and you can approve
companies everyone else is turning away.
What version 1 actually was
They dropped the VR idea completely and pivoted inside that same batch to a corporate
card for startups, underwritten against the company's bank balance rather than a personal
guarantee. It was the second payments company either founder had built. They had already
sold an earlier one, Pagar.me, to StoneCo in 2016.
How it makes money
Interchange on the corporate and spend-management cards, plus software revenue from the
expense and finance platform. Cards, spend tooling and banking are bundled into one
product rather than charged for separately.
Where it is now
Capital One announced a definitive agreement to acquire the company on January 22, 2026,
for $5.15B, roughly $2.75B in cash plus about 10.6 million Capital One shares, well under
half the October 2021 peak private valuation of $12.3B. That deal closed April 7, 2026,
and Brex now operates as part of Capital One with Pedro Franceschi continuing as chief
executive. At announcement it had roughly 1,100 employees, was growing about 40% year over
year, and was profitable.
The student-sized version
- Problemyou are too new to have the track record a gatekeeper wants, even though you are good for the money.
- Useryourself, at the age you are right now, trying to get approved for something that assumes history you have not had time to build.
- V1find a different signal to underwrite against instead of the one the incumbent insists on.
That is a Rocky Top entry: started at your age, on a wall the founders had just hit themselves.