On 22 May 2010, a Florida programmer named Laszlo Hanyecz paid 10,000 Bitcoin for two pizzas. At the time that was about forty-one dollars. At the time of writing, those same coins are worth several hundred million. The story has a name now, Bitcoin Pizza Day, and it gets reshared every May with the kind of affectionate horror reserved for people who did a very reasonable thing that turned out to be catastrophically expensive.
Fifteen years on, that one transaction has done more to stall crypto commerce than any regulation or engineering problem. It sits in the back of every crypto holder’s head as a tax on spending, and the tax is psychological.
The standoff
Survey after survey tells the same story. Around 55% of crypto holders rarely or never spend their holdings, even though roughly 80% of those same holders say they believe in broader adoption. The single biggest reason they give is fear of missing future appreciation. Fear of becoming the Pizza Guy. Regulators looking at the same population reach the same conclusion: most retail holders treat crypto as a long-term investment, not a means of payment.
On the other side of the counter, merchants who accept crypto almost always convert it to fiat the moment it lands, because they can’t afford to hold an asset that might fall 20% before the week is out. Payment gateways like BitPay and CoinGate built a good business out of this conversion. The merchant side of the problem has been solved for years.
The hard part is the holder. A person with crypto they like, standing in front of a thing they want, deciding not to spend. Multiply that by fifty million wallets and you get roughly a trillion dollars of asset value generating less than one percent of its potential transaction volume.
Why the usual workarounds don’t work
There are two informal ways people solve this today, and both reveal something interesting about why a cleaner fix hasn’t shipped.
The first is spend and replace. Apps like Strike and Fold let you spend crypto, and then immediately rebuy the same notional amount from your linked bank account. Your crypto balance stays constant. But you need fiat sitting idle in your bank to do it, you pay fees on both legs, and the approach only works if you happen to have the cash. Most people don’t.
The second is borrow, don’t spend. You pledge your crypto as collateral, borrow stablecoins against it, and buy the thing with the loan. Your crypto never moves, you keep the upside, and you’ve converted a spending problem into a loan-to-value management problem. Which is a real problem. If the price drops enough you get liquidated, which is exactly the scenario the user was trying to avoid.
Both workarounds shift the problem around. Neither of them say “yes, spend, and if the price runs you’ll be looked after.” That sentence is what people actually want, and no one is allowed to say it.
Why no one says it
The short answer is that saying it properly means selling financial protection to retail buyers, and selling financial protection to retail buyers triggers a regulatory gravity well most startups are not equipped to survive.
We spent months mapping every shape the product could take. A consumer brand. A firm that would take the risk onto its own balance sheet. A cashback product that looks like upside protection by coincidence. A sweepstakes that pools premiums into a prize pot. Every shape we tried either required a licence that would take years to get, priced normal users out, or collapsed the moment a regulator looked at it carefully.
The shape that survived was narrower than any of those. The thing people want can be offered legally, but not by a startup. It can be offered by a regulated platform that already holds customer funds, already sits inside a compliance wrapper, and already has a venue account big enough to hedge real volume. What a platform like that doesn’t have is the infrastructure to price the protection, lock the terms, and produce the evidence trail on every transaction at retail scale without spinning up a dedicated desk.
That infrastructure is the piece we build.
We call it the Calculation Agent. It is software. It prices a bounded protection contract, records the decision, tells the partner what to hedge, and hands the finance team a clean settlement trail. The partner, who is already a regulated entity, keeps the customer, keeps the rails, and keeps signing authority on the hedge.
Why we’re building it now
Two things changed.
The first is the size of the card programmes. Coinbase, Crypto.com, Binance, and the other large crypto platforms have spent years building debit cards that convert crypto to fiat at point of sale. The merchant never sees crypto. The card issuer earns interchange. These programmes already process meaningful volume, but the 55% of holders who never swipe are billions of dollars of interchange left on the table. The conversations we have had with these programmes are direct. Nobody disputes the opportunity. Nobody disputes that the fix has to live one layer up.
The second is that autonomous agents now spend money. Within the next few years a meaningful share of transaction volume online will come from software agents paying for inference, tools, data, and services on behalf of a human principal. An agent does not feel FOMO, but it can absolutely bankrupt a treasury by mistiming a payment during a flash crash, and the humans who authorised the agent are going to want a demonstrable risk floor underneath every decision the agent makes. The HTTP standard emerging for machine payments (x402) gives a clean interception point to add that floor without making every agent developer invent their own.
So we built one engine with two ways in. A standard API for human checkout flows. An interceptor for agent payments. The logic underneath is the same: price the downside, lock the terms, let the partner execute, hand back an evidence trail.
What Slice is not
It helps to name the shapes we chose not to be.
We are not an insurer. We do not carry customer risk and we do not pay claims out of our own balance sheet. The partner does that, under their existing regulatory wrapper.
We are not a wallet. We do not hold your crypto, we do not custody anything, and we have no interest in being in front of the customer. The wallet you already use is the right place for the customer relationship.
We are not an exchange. We do not route orders or match trades. The partner’s existing venue accounts do that.
We are software. The unglamorous, legible kind. The value sits in one API call that returns a locked, bounded quote, a stream of settlement events, and a reporting surface that a finance team does not have to decrypt.
What a good pilot looks like
The metric that matters on a pilot is conversion uplift. If protected checkout converts meaningfully higher than unprotected checkout on a real commerce surface, the business case writes itself. That is the A/B test every first pilot is designed around.
If that sounds like a conversation you would like to have, we would love a fifteen-minute call. Bring a flow.