Product Engineer
JavaSpring Boot
Role Snapshot
Hands-on Java experience, an end-to-end shipped product, experience with agentic coding tools, strong skepticism when reviewing AI output, and intermediate to advanced written and spoken English.
Job description
AstroPay holds the licenses. AP Platform is what we build on top of them: infrastructure that lets any company launch its own financial products without becoming a financial institution first. An airline, a marketplace, a retailer. Wallets, payments, checkout. Our partners are all over the world, and real money moves through what we ship.Our team is smaller than it was a year ago. Instead of putting it back the way it was, we're rebuilding it around different principles.
Job requirements
We're leaning hard into AI across the whole development process and paying close attention to where it helps and where it doesn't. Much of that is still being worked out, by us and by everyone else. We'd rather hire people who want to be part of working it out than people waiting for the finished version.You won't be handed one service to look after. You'll move across backend, frontend, infrastructure and data, sit in the product conversations, and get close enough to the business to know why a partner is asking for something before you build it. Everyone here talks to those partners directly, not only the leads.
AI writes a large share of the code. What stays human is the good part: deciding what to build, knowing whether what came back is right, and taking it into production yourself.
Our core is Java and Spring Boot, and it will be for a long time. You don't need to know it deeply, but you need enough hands-on time to read and write it comfortably while you pick up the rest.
Job responsibilities
How you work matters more to us than how long you've been working. Early in your career, or fifteen years in and ready to rebuild how you build: both work here.• Hands-on experience with Java. Spring Boot is a plus, not a requirement.
• Something you built end to end and shipped. What matters is that people used it.
• A working relationship with agentic coding tools that goes past autocomplete. You should have opinions about what they're bad at.
• Skepticism. You read what the model produced before you trust it, and you've caught it being fluent and completely wrong. That instinct is the one thing we can't teach quickly.
• Written and spoken English, intermediate to advanced.
Plenty of people are waiting for this shift to pass over them. We're looking for the ones who'd rather be on top of it.