Software Development · Berlin, Germany

Senior Software Engineer, Agentic Platform

GetYourGuide · Berlin, Germany · Posted Oct 19, 2026

This posting comes from GetYourGuide. Confirm the details on their site before applying.

Overview

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.

Responsibilities

  • 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.
  • 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.
  • 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 that 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 your success metrics.
  • Annual personal growth budget and mentorship programs for continuous learning and development

Benefits

  • Wellness program

Great fits start here.

Whether you're looking for your next role or your next team member, give people and possibilities a place to meet.