"As 2025 ends, we’re recounting this year’s launch, the Request Network API allowing builders to accept crypto payments, manage payouts, and invoicing flows via API endpoints, with many others to come next year."
I personally think payment tech should stay as in-house as possible for security reasons. Can't something be vibe-coded? Don't give others the temptation to spy.
And from my understanding and slight observations, the French are experts in commercial espionage. Please, someone disagree.
BUT, for you this is a French startup, so I thought you'd like it.
Good point on smart contracts: that's one approach I'm exploring.
Trying to figure out if this should be:
- Crypto-native from day 1 (smart contracts + stablecoins)
- Fiat-first with crypto as backend (abstract the blockchain)
- Pure fiat (traditional banking APIs)
Would be interesting to have a way for each recipient to opt-in for their preferred currency. Am thinking about recent headline: "Crypto payment cards surge 22x in daily transactions since late 2024."
Just curious, if you didn't have smart contracts, you would have 20 destination addresses, and then a spreadsheet that you'd use each month and then run a local wallet to do the sending to the different parties, right? That's pretty easy.
Could that be automated on the client-side with macros or a custom program (ie no smart contracts?)
Multisig with no smart contract is the simplest escrow.
It's probably simple to find a validated smart contract that splits its inputs to multiple outputs?
If there is transaction privacy for the smart contracts, then how to verify that the contract actually still sends to the correct - initially configured - parties?
And then logging. Task accounting implies logging, but logs are bytes that cost money in a blockchain so is the transaction log sufficient
Aside from the fact that it's very funny to see Mistral and others performing Hadouken, we found that it is a great way to benchmark language models. They need to quickly understand their environment and take actions accordingly.
New Relic is close to Datadog, more focused on observability for engineers. We're focusing on extracting insights from the text. We focus on NLP more than trace analytics
I'm looking for: - Dynamic rules based on revenue amount - Multiple variable destinations - Percentage-based splits - One-time setup, runs forever
Maybe I'm missing a Mercury feature? What's the workflow you use?
Im a Brex user and interested by this