> Markdown version of [/jobs/ext/4780322-principal-senior-software-engineer-lightwell-ecosystem-spain](https://www.wearedevelopers.com/jobs/ext/4780322-principal-senior-software-engineer-lightwell-ecosystem-spain). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Principal/Senior/Software Engineer - Lightwell Ecosystem (Spain) - **Company:** Red Hat - **Location:** Madrid, Spain (Remote available) - **Contract:** Permanent contract - **Skills:** Artificial Intelligence, Continuous Integration, Software Debugging, Machine Learning, Software Construction, Systems Integration, Prompt Engineering, Build Tools, Software Version Control - **Published:** November 19, 2026 - **Apply:** https://dejobs.org/x/x/5E294864C15F43C597F186B11BA76748/job/ ## About the Role * Experience building software, automation, or analytical tooling that others use and maintain. * Strong debugging and problem-solving skills, including the ability to investigate failures across system boundaries. * A habit of testing claims against evidence and improving systems based on observed outcomes. * Clear communication and a collaborative approach to documenting decisions, results, and limitations. ## Description * Design, build, and maintain the agentic pipeline that generates candidate patches - orchestration logic, model integration, prompt engineering, and tool-calling that make automated remediation reliable at scale * Integrate AI/ML models and agentic frameworks into pipeline workflows: routing, retries, guardrails, and evaluation loops that improve patch quality and success rate over time * Build and maintain the CI/CD, testing, and validation infrastructure that supports high-volume automated remediation work across ecosystem teamsç * Troubleshoot and resolve pipeline failures, flaky automation, and integration issues spanning source control, build systems, and downstream handoffs * Instrument and monitor pipeline health - success rates, false positive rates, latency, failure classes - and drive continuous improvement from that data rather than anecdote * Partner with Language Ecosystem Engineers to resolve pipeline fix tickets raised when capability gaps block their work; turn one-off blockers into durable pipeline improvements * Evaluate and adopt new automation frameworks, agentic tooling, and AI capabilities that could improve pipeline throughput, reliability, or quality * Document pipeline architecture, runbooks, and known limitations so the system is operable and auditable by people who didn't build it