ph3About Orbio /h3 pOrbio is an AI-native human capital management platform building a bDigital Workforce for HR /b. /p pWe design and deploy agentic systems that take ownership of operational execution across hiring, onboarding, and employee operations — while keeping humans in control. /p pWe work with large, complex organizations across staffing, retail, hospitality, and healthcare. Our customers rely on Orbio to reduce operational friction, improve speed and quality, and scale their teams without proportional headcount growth. /p pOrbio is not a system of record. We sit on top of existing HR systems and own the bexecution and coordination layer /b — where day-to-day work actually happens. /p h3Position Overview /h3 pOrbio is looking for an bAI Agent Engineer /b to build, operate, and continuously improve the reliability of our agentic AI systems in production. /p pIn this role, you will focus on how agents bbehave, decide, fail, recover, and scale /b in real enterprise environments. Your work will ensure that LLM-powered agents are observable, testable, cost-efficient, and trustworthy when handling real operational workflows. /p pYou will help build an end-to-end, production-grade B2B platform with a strong focus on LLMs, autonomous agents, and conversational AI, contributing to the technical foundations that scale across customers and use cases. /p h3All the desired Technical Skills we are looking for /h3 pbAgent Testing Evaluation /b /p ul liBuilding agent evaluation pipelines (consistency, accuracy, reliability) /li liDesigning test harnesses for non-deterministic systems /li liRegression testing for prompt and model changes /li liSynthetic data generation for agent testing /li liPerformance benchmarking and latency profiling /li /ul pbAgent Monitoring Operations at Scale /b /p ul liObservability for agent systems (tracing, logging, metrics) /li liCost tracking and optimization across model calls /li liAlerting on agent failures,



hallucinations, and degraded performance /li liDebugging complex multi-step agent behaviors in production /li liGuardrails, safety layers, and graceful degradation strategies /li /ul pbLLM Expertise /b /p ul liPrompt engineering, chaining, and caching strategies /li liRAG pipelines and vector databases /li liModel selection, fine-tuning, and cost optimization /li liUnderstanding of evaluation metrics and benchmarks /li /ul pbPython Backend /b /p ul liAsync programming, Django/Channels, real-time ASGI /li liProduction-grade application development /li liMessage queues (Celery + Redis) and high-throughput systems /li liTesting, logging, and performance optimization /li /ul pbConversational AI /b /p ul liVoice pipelines: STT, TTS, turn-taking, interruption handling /li liTelephony integration (Twilio) with latency optimization /li liConversation flow design and fallback strategies /li /ul pbFull-Stack DevOps /b /p ul liFrontend experience (React preferred) /li liCI/CD pipelines and cloud deployment (AWS/GCP) /li liDeploying and operating AI systems in production /li /ul h3What Success Looks Like at Orbio /h3 ul liAgents behave consistently and predictably across customers and workflows /li liFailures, edge cases, and degraded behavior are detectable and recoverable /li liAgent performance, latency, and cost are measurable and actively optimized /li liNew agent capabilities ship safely without breaking existing behavior /liliCustomer trust in autonomous workflows increases over time /li /ul h3Culture at Orbio /h3 pOrbio is a high-bar, high-intensity environment. We optimize for impact, speed, and ownership — not comfort. /p pWe operate as one team, with a flat structure and extreme ownership. We ship fast, learn from production, and challenge each other directly because we care about building systems that hold up under real-world complexity. Titles matter less than outcomes. Customers come first. Momentum matters. /p /p #J-18808-Ljbffr

📌 AI Agent Engineer (Python) / Full-Stack experience (España)
🏢 Orbio
📍 España