Full Stack · Payments squad · LLM-first by default.
Join our Payments squad to build out a cross-market flexible Payments capability — at a critical moment for how the team works. We've recently transitioned to LLM-first engineering, so you'd help shape both our ways of working and our approaches to LLM-based tooling. Polyglot stack, full ownership end-to-end, and a high bar for engineering quality.
See the roleMore than anything we're interested in your general experience as a software engineer and your approach to solving problems. We're hiring at an exceptional bar — not just strong code, but real ownership, judgment, and impact beyond your own keyboard.
Claude Code, Codex, Cursor are part of how you build — not afterthoughts. You can show real iteration with these tools, push back on their output, and explain decisions the AI made on your behalf. You're motivated to keep pushing as agentic LLMs reshape what good engineering looks like.
5+ years in full-stack engineering, comfortable across TypeScript, Python, or Ruby. Familiar with web frameworks like FastAPI or Ruby on Rails. Specific implementation language isn't critical — you adapt to what the problem needs.
Self-directed: takes a vague outcome and turns it into shipped work. Builds consensus across teams, makes decisions without waiting to be told, and raises the bar for the engineers around you. Excited by Ophelos's mission.
You'll join the Payments squad to build out our cross-market flexible Payments capability. The team has recently transitioned to an LLM-first working style — you're joining at a critical moment to shape both how the squad works and how we use LLM-based tooling. Full ownership end-to-end, with meaningful and measurable impact on the org and our products.
Most days are a mix of design, build, and review — almost all of the build done through LLM-driven workflows (Claude Code, Codex, OpenSpec) rather than direct edits. Specs and designs are drafted with AI before any code is written. Three days a week you're in our Liverpool Street office with the Payments squad for planning, alignment, and the kind of conversations that don't happen on Slack; the other two are heads-down. You'll regularly be deciding architecture for features that touch how money moves, and pushing the team's LLM workflows forward at the same time.
Every company says they use AI. The real question: if you removed it, would the product still work? At Ophelos, the answer is no — the core of what we do (deciding who to contact, how, and when) leverages AI systems and machine learning. It's not a rules engine with a chatbot bolted on.
The stack is Claude-native end to end. Engineers use Claude Code and VS Code as their primary environment, with AI embedded at every stage — writing, reviewing, debugging, refactoring. The product itself runs on the same stack: a purpose-built AgentOS powered by Agno, routing to Claude and OpenAI through an internal LLM gateway. Specs and designs are drafted with AI before code; engineers use a structured workflow where AI helps write proposals, technical designs, and task lists — it's a partner on architecture, not just a code completer.
Four stages, designed to be a real conversation rather than a series of hoops. We aim to come back within 48 hours after each round.
