About this role
1 Principal Software Engineer (m/f/x) - Hyperscaler & Strategic Partner Integrations (AI)
What you will do
- Design and build integrations between Dynatrace and hyperscaler AI / observability services
- Build and maintain Dynatrace's Model Context Protocol (MCP) servers so AI assistants and agents can query problems, traces, logs, and Dynatrace Intelligence insights as first-class tools - and contribute upstream to the MCP spec where it matters.
- Co-develop with IDE providers (Anthropic / Claude Code, GitHub Copilot, JetBrains AI, Cursor, Windsurf, AWS Kiro, Gemini CLI) so Dynatrace observability shows up where developers already work - plugins, extensions, slash commands, MCP integrations.
- Co-develop with partner engineering teams at AWS, Microsoft, Google, and IDE vendors - joint design docs, shared backlogs, paired debugging across org boundaries.
- Extend Dynatrace's OneAgent and OpenTelemetry pipelines to capture traces, metrics, and evaluations from agentic AI workloads (tool calls, prompt chains, model latency, hallucination signals).
- Contribute to launch-partner integrations - being one of the first observability providers shipping against a new hyperscaler or IDE-vendor service requires fast iteration against pre-GA APIs.
- Own a slice of the partner-facing SDK, MCP server, or IDE-extension surface end-to-end: code, tests, docs, marketplace listing, on-call.
- Represent Dynatrace engineering in technical reviews with partners; translate Dynatrace Intelligence and causal-AI capabilities into terms partner engineers care about.
- Mentor other engineers and raise the bar on how the team works across org boundaries.
What will help you succeedEngineering foundation
- 5+ years of professional software engineering experience building and operating production backend systems.
- Strong proficiency in at least one of Java, Go, Python, or TypeScript; able to ramp on a second within a quarter.
- Designed and shipped public-facing or partner-facing APIs (REST and / or gRPC), including versioning, backwards compatibility, and SDK ergonomics. Hyperscaler depth (at least one)
- Hands-on, production-grade experience with AWS, Azure, or GCP - building against native SDKs, not just deploying to VMs.
- Concrete familiarity with the partner-relevant surface of that cloud:
- AWS: Bedrock / AgentCore, Lambda, EKS, CloudWatch, AWS Marketplace, Partner Network (APN).
- Azure: Azure OpenAI, AKS, Azure Monitor, SRE Agent, Azure Marketplace, MISA.
- GCP: Vertex AI / Gemini, GKE, Cloud Operations, Google Cloud Marketplace, Partner Advantage. Observability and AI workloads
- Working knowledge of observability primitives: metrics, logs, distributed tracing, OpenTelemetry instrumentation.
- Understanding of how to instrument LLM and agentic-AI workloads: prompt / response capture, token usage, tool-call traces, evaluation signals, latency and cost attribution.
- Comfortable reasoning about SLOs, error budgets, and incident response in customer-facing systems.MCP and IDE-provider integration
- Practical experience with the Model Context Protocol (MCP) - building MCP servers, clients, or both. Comfortable with stdio and HTTP transports, tool / resource / prompt primitives, auth, and the realities of running MCP in production.
- Built at least one IDE extension, plugin, or integration that real developers used - for VS Code, JetBrains, Cursor, Claude Code, GitHub Copilot, Gemini CLI, Kiro, or similar.
- Understands the AI-assisted developer workflow end-to-end: how agents discover tools, how context windows fill up, how tool outputs get re-ranked, why a well-designed tool schema matters more than a clever prompt.
- Can reason about trade-offs of agent-facing API design vs. human-facing API design (token cost, determinism, error surfacing, idempotency).Compensation and RewardsDue to legal reasons, we are obliged to list a salary range for this position, which is € 90,000 - € 112,000 gross per year based on ...