Descripción del trabajo

Experteer Overview As a senior architect, you will design AI-driven solutions for ITSM and AIOps, focusing on generative AI, agents, RAG, and automation. You will define architectures for agents, tools, memory, and decision flows, and guide the implementation across operations, support, and service management. You'll shape prompts, evidence retrieval, scoring, and human-in-the-loop strategies to ensure safe, scalable automation. You'll collaborate with technical teams and stakeholders to deliver impactful, enterprise-grade AI solutions that continually improve. This role offers a hands-on, strategic opportunity to drive automation at scale within a modern tech stack.

Compensaciones / Ventajas

Design generative AI solutions for business and operational processes

Define agentic architectures including agents, subagents, tools, memory, context, workflows and control points

Model AI flows for ITSM/AIOps processes (classification, prioritization, assignment, diagnosis, resolution, scaling, closing, reopening)

Define RAG patterns across documentation,



knowledge bases, tickets, runbooks, logs, metrics and traces

Develop prompting strategies, structured generation, evidence retrieval and response evaluation

Create scoring models (trust, similarity, risk, criticity, automation eligibility, response quality)

Design human-in-the-loop mechanisms, policy gates, and criteria for assisted/supervised/autonomous automation

Integrate AI agents with APIs, internal systems, cloud platforms, ITSM tools, observability tools, external services and corporate data sources

Identify automation candidates from process analysis and historical data

Participate in automation service development and automatic code generation

Coordinate and guide technical teams on AI solutions using agents, RAG, workflows and automation

Drive continuous improvement of agents, prompts, workflows, tools and evaluation mechanisms

Stay updated with advances in generative AI, LL

📌 Agentic Ai (Madrid)
🏢 Deutsche Telekom
📍 Madrid