Senior AI Agent Engineer (Go/Python) - Build LLM Systems

Dynatrace; Oberösterreich; Vor Ort; EUR 61.000 - 76.000; Vollzeit

Benefits dieser Stelle

Relocation support to Linz

Zusammenfassung

Dynatrace is seeking a practical software engineer to turn LLM reasoning into safe, reliable agentic behaviors in production. You will design and maintain the Python agent layer and guardrails, while improving evaluation strategies for agent quality across model updates.

Successful candidates bring deep Go and Python experience, hands-on work with LLM APIs and agent loops, and a proven ability to measure system quality using quantitative data. Relocation support to Linz is provided.

Qualifikationen

• Multiple years of professional, hands‑on software engineering experience with deep expertise in Go and Python.
• Direct experience building LLM-powered systems and agent loops, utilizing modern orchestration frameworks (like LangGraph), tool-calling, and structured-output APIs.
• Demonstrated ability to measure, evaluate, and prove agentic system quality through quantitative datasets and evaluation metrics rather than qualitative impressions.
• Strong software engineering fundamentals, including continuous testing, production debugging, and full-service ownership.
• Degree in Software Engineering or equivalent practical experience in software development.
• Highly developed analytical, problem-solving, and collaborative skills to work effectively within an international engineering team.

Aufgaben

• Design, build, and maintain our core Python agent layer, including LLM routing, tool-calling infrastructure, and structured-output retry logic that ensures dependability in production.
• Design robust guardrails, input and output validation, and prompt-injection defences to mitigate hallucinations and prevent agents from deviating from their tasks.
• Define and implement the datasets, metrics, and automated evaluation strategies necessary to measure and prove agent quality objectively across model updates and prompt modifications.
• Drive high-level architectural decisions for our AI systems and actively participate in team code reviews, unit testing, and debugging efforts.

Kenntnisse

Go

Python

LLM orchestration

Agent loops

Production debugging

Analytical mindset

Ausbildung

Degree in Software Engineering

Tools

LangGraph

LLM tooling

Jobbeschreibung

Dynatrace is seeking a practical software engineer to turn LLM reasoning into safe, reliable agentic behaviors in production. You will design and maintain the Python agent layer and guardrails, while improving evaluation strategies for agent quality across model updates.

Successful candidates bring deep Go and Python experience, hands-on work with LLM APIs and agent loops, and a proven ability to measure system quality using quantitative data. Relocation support to Linz is provided.