AI Software Engineer

About the role

We are building the way FTMO ships software in the age of AI, and we want you to own it. This is not a "use a bit of Copilot" role. We are making AI-first delivery the default across our IT department, and you will be the engineer who makes that real.

You will own our AI development harness: the coding-agent patterns, conventions, and end-to-end loops that let agents make reliable, incremental progress across long sessions and many context windows. Think of it as building the AI software factory the whole company builds on top of.

This is a hands-on senior role for someone who already ships production-grade software primarily by directing coding agents, not by hand-writing every line. You set the example, the guardrails, and the conventions that every other engineer's agents will follow.

WHAT WILL BE YOUR AGENDA?

- Design and own FTMO's AI development harness: initializer and coding-agent patterns, AGENTS.md and CLAUDE.md conventions, progress and checkpoint files, feature lists, and end-to-end test loops.

- Shape the AI-native SDLC and PDLC together with our DevEx Specialist. Map every phase to a clear delegate / review / own model, then wire the tooling (MCP integrations, eval loops, CI hooks).

- Run fast proofs of concept against real codebases and workflows, then turn the ones that hold up into production-ready patterns.

- Deliver production software agent-first, keeping the codebase clean and mergeable and setting the standard for agent-first delivery.

- Build and harden the AI infrastructure: model access and routing, sandboxes, secrets and permissioning, observability, and cost controls for agent runs.

- Raise the whole department's AI fluency through patterns, internal documentation, and hands-on pairing, and help non-engineering teams ship and govern their own apps.

- Track the frontier. Evaluate new models and tooling, and bring the ones that earn their place into our stack, including reporting.

WHAT DO YOU BRING TO THE TABLE?

- You are a senior SW engineer who has shipped and owned production software across the full lifecycle.

- Around a year working AI-first: delivering real, production-grade software primarily by directing coding agents.

- You can show us the harness. Agent configs, AGENTS.md and CLAUDE.md files, eval suites, prompts, MCP integrations, or repositories that prove how you get agents to do reliable work.

- Strong general software engineering. You can read, design, and review production code in at least one major stack and reason about architecture, testing, and security.

- Deep, practical fluency with agentic coding tools (Claude Code and the Claude Agent SDK, Codex, Cursor, or equivalents): context management, compaction, long-running sessions, and multi-agent patterns.

- Prompt and context engineering as a daily discipline, plus a feel for evals and verification loops so agent output is measured against tests, latency, and style