Role overview
Join TDC NET and help us build and scale advanced agentic systems. At TDC NET, we are leaders in Denmark's digital infrastructure, dedicated to technical excellence, sustainability, and continuous improvement in an ever-evolving industry. Our team is developing robust LLM systems that drive internal automation and decision support, making a significant impact within the organization.
About the job
You will be the specialist on LLM architecture, agent design, and best practices, directly involved with coding and setting the technical direction for the company's LLM initiatives. Leading the design and development of agentic workflows, you will make key decisions on models, orchestration, and infrastructure, while establishing standards for reliability, security, and performance. Expect to engage in prototyping, pair programming, and meticulous engineering, helping other teams efficiently deliver solutions.
As a crucial member of an innovative product team—including a Product Manager, Platform Engineer, and LLM developers with GenAI experience—you'll create a collaborative environment focused on collective learning, creating valuable user-centric products, and enjoying the process.
What you'll work on
You will take charge of building production-grade LLM systems that empower internal automation and decision support:
- Design and implement agentic systems including planning/execution loops, tool usage, memory, and multi-agent patterns.
- Be highly skilled in LLM orchestration encompassing LangGraph/LangChain/MCP, routing strategies, function/tool calling, and safe tool integration.
- Build solid RAG pipelines focusing on indexing strategies, retrieval quality, evaluations, and latency/cost optimizations.
- Integrate with our stack through APIs, internal services, databases, and event-driven workflows, establishing clear contracts and service level agreements (SLAs).
- Work with the processes including containerization, Kubernetes deployment, GPU sizing, autoscaling, and observability (metrics, traces, logs).
- Establish GenAI engineering standards like prompt/version management, offline/online evaluations, AB testing, guardrails/safety, and incident playbooks.
- Drive review by setting patterns, running design reviews, and contributing reusable components.
- Stay current and pragmatic by assessing new techniques and adopting those that deliver measurable value.
"Our LLM agents are built to make real decisions easier for real people. You'll work close to the problem, shape the solution, and see your work land across the company to support Denmark’s digital backbone", says Jesper Wass, Director, AI Application Engineering, TDC NET.
About you
*Who we are looking for
Must-haves:**
- Collaborative colleague, motivated to share knowledge and explore the unknown.
- 1+ year building and shipping LLM-powered systems (RAG, workflows, agents) to production and 3+ years in software/backend/ML
- LangGraph/LangChain, MCP, or similar frameworks in production.
- Skilled in Python and engineering practices (Gitlab CI/CD, testing, observability).
- Experience with LLM orchestration concepts: tool/function calling, workflow graphs, memory/state, retries/timeouts, and evals.
- RAG in practice: indexing strategies, retrieval quality, latency/cost tuning, and offline/online evaluation.
- Reliability mindset: tracing/metrics/logging, performance tuning, incident response, and guardrails/safety.