Senior Agentic Engineer

Elestio is hiring a Senior Agentic Engineer\n\nThe way we build software is changing fast, and we're looking for someone who's excited to build at the heart of that shift.\n\nElestio is the managed platform for open-source software: 400+ tools deployed and operated across 9+ cloud providers, with European data sovereignty, SOC 2 and ISO 27001. We run real infrastructure for real companies, so quality and reliability are at the center of everything we do.\n\nWhat you'll do:\nOrchestrate AI agents to design and ship features at a pace solo coding can't match\nBuild the verification systems we rely on: tests as the real contract, not just a quality ritual\nRun dedicated security passes for the way AI code tends to fail: permissive defaults, \"just in case\" fallbacks, loose secrets handling\nDesign invariants and production monitoring for the issues that only show up under real traffic\nOwn our agentic and MCP infrastructure end to end\n\nWhat we're looking for:\nAt least 5 years of hands-on full-stack development experience\nSolid command of our stack: React, Node.js, SQL, Bash and Linux\nStrong adversarial thinking: you naturally ask \"what would break this?\"\nSolid test design and healthy, calibrated distrust of generated code\nSenior fundamentals: you have a good sense of where AI tends to cut corners\nComfort with agents, LLM tool-use, orchestration and MCP\nA bias for shipping and for owning outcomes\n\nThe skills that matter most here are adversarial thinking, test design and the ability to look at a feature an AI built and tell whether it's safe to ship. In fact, that's close to how we interview: we'll show you a feature an agent wrote, and we'll explore together how you'd make sure it's ready for production.\n\nWhy you'll enjoy it:\nReal ownership and room to shape the role\nProduction from day one, on infrastructure that matters\nOpen-source culture and compliance-grade engineering, together\nA small, senior team that ships fast and respects your time\n\nIf you've been building with agents and you think in tests, invariants and failure modes, we'd love to talk.