Senior Software Engineer, Agentic Platform
Senior engineer building GetYourGuide's internal agentic platform (Midnight Clue): designing core features and tooling, owning the LLM evaluation/observability layer (tracing, evals, feedback loops), and shaping end-to-end developer experience so teams can delegate work to agents. Core stack: TypeScript, Go, Kubernetes, LLMs in production.
- Compensation
- Not disclosed
- Location
- Berlin
- Work setup
- Hybrid
- Seniority
- Senior
- Employment
- Not specified
- Experience
- 6+ years
- Category
- Software engineering
- Posted
Skills
Must have
- 6+ years of software engineering experience
- Track record building and operating platform-level or developer-facing systems other engineers depend on
- Strong hands-on skills in TypeScript and Go, with judgment on when to use each
- Solid foundations in distributed systems, APIs, and cloud-native deployment (Kubernetes or similar)
- Practical experience building with LLMs in production: prompt and tool orchestration, retrieval, agent loops, and cost/latency/reliability trade-offs
- Measurement mindset: built or run eval pipelines, tracing, or monitoring for systems
- Ability to lead technically without a title: clear written and verbal communication, cross-team influence, mentoring
Nice to have
- Experience with LLM observability or evaluation tooling (e.g. Datadog LLM Observability, LangSmith, Braintrust, OpenTelemetry-based tracing)
- Familiarity with agent protocols and frameworks such as MCP, and with building SDKs or CLIs other developers adopt
- Background in security, guardrails, or policy enforcement for AI systems (prompt injection defense, PII handling, permissioning)
- Prior work on an internal platform or DevEx team where adoption and developer satisfaction were success metrics
Full description
Change the way the world travels
Join the GetYourGuide journey to connect people with unforgettable travel experiences around the world. Millions look to us for unique activities they can trust, and it’s all powered by our commitment to make every single journey extraordinary - including yours.
Ready to unlock your potential with a community of fellow explorers? Find your next role at our Berlin HQ or one of our local offices around the globe, from New York to Bangkok. Head to to take the first step.
Team mission
We build the agentic platform that GetYourGuide engineers use to hand off toil. Engineers bring the problem and the outcome they want. The platform already knows how GetYourGuide works and handles the plumbing, so teams can focus on business impact.
Your mission
- Design and build the core features and tooling of GetYourGuide's internal Agent Platform, Midnight Clue, so product and engineering teams across the company can define their context and delegate work to the agent.
- Own the evaluation and observability layer: tracing, quality metrics, offline and online evals, and the feedback loops that tell teams whether their agents are getting better or worse in production.
- Shape the user experience end-to-end, from how a team defines the context to how they test, monitor, and optimize their agentic workflows, with guardrails baked in by default.
- Drive architectural decisions for the platform, write RFCs, and partner with infrastructure and application teams to ensure what we build fits GYG's broader systems.
- Raise the bar for the team through code review, mentoring, and pairing, and act as a go-to technical voice for agent development across the organization.
Your toolkit
- 6+ years of software engineering experience, with a track record of building and operating platform-level or developer-facing systems that other engineers depend on.
- Strong hands-on skills in TypeScript and Go, with sound judgment about when to use each, and solid foundations in distributed systems, APIs, and cloud-native deployment (Kubernetes or similar).
- Practical experience building with LLMs in production: prompt and tool orchestration, retrieval, agent loops, and the cost, latency and reliability trade-offs that come with them.
- A measurement mindset: you have built or run eval pipelines, tracing or monitoring for systems and can explain what "good" looks like with data rather than vibes.
- The ability to lead technically without a title: you communicate clearly in writing and in person, influence decisions across teams, and help others level up.