> Markdown version of [/jobs/ext/3754363-software-engineer-i-ai-observability](https://www.wearedevelopers.com/jobs/ext/3754363-software-engineer-i-ai-observability). 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).

---

# Software Engineer I - Ai Observability

- **Company:** Roleelastic Observability
- **Location:** Madrid, Spain
- **Contract:** Permanent contract
- **Skills:** Artificial Intelligence, Software Code Optimization, Software Debugging, TypeScript, ReactJS, Multi-Agent Systems, Software Application Programming, Generative AI, Information Technology, Machine Learning Operations
- **Published:** October 15, 2026
- **Apply:** https://www.adzuna.es/details/5928230637

## About the Role

Help build offline and online evaluations, and contribute to dataset construction and curationCross-functional collaboration: Partner with UX designers on how autonomous investigations are presented, product managers on scope and sequencing, and data scientists on analysis methodsWhat You BringAt least two years developing applications with TypeScript, React and Node.jsUsing AI to accelerate development, debug systems, and optimize code, while still owning the outcomesKnowledge of LLM-based systems - agent orchestration, tool use, retrieval, context managementA bachelor's degree in Computer Science (or a related technical field), or equivalent hands-on experienceA creative individual with an enthusiastic interest in technology and software engineeringA willingness to learn in a dynamic and distributed environmentBonus PointsKnowledge of self-directed agentic execution (multi-turn reasoning loops) and evaluating output qualityEvaluating non-deterministic systems by
designing evaluation datasets, offline and online evaluationsKnowledge of Observability and ElasticsearchExperience working in distributed teamsExperience in open sourceCompensation for this role is in the form of base salary.

## Description

This is most of the work today, though investigations increasingly need to reach engineers wherever they already workAgentic investigation: Contribute to the multi-agent system that pursues root cause autonomously: build the tools agents use to investigate, shape the context they reason over, and help decide when an agent has gathered enough evidence to concludeDetection and signal generation: Contribute to the systems that establish baselines across high-cardinality telemetry, generate their own detection queries, and identify state changes worth investigatingEvaluation infrastructure: Agent output quality is the product here.