4 hours ago
Berlin, GermanyStaff+
Responsibilities
- Build agent orchestration, context management, memory, and programmatic tool-calling systems.
- Enable agents to perform useful work through sandboxed code execution and permission-aware data access.
- Develop evaluations for multi-turn workflows to measure agent quality and reliability.
- Integrate models and build inference infrastructure for practical AI operation at scale.
- Create shared foundations that help product engineers deliver agent-powered features.
- Own LLM-powered systems after launch and help other engineers succeed.
Requirements
- Typically 5+ years of software engineering experience.
- Strong software engineering fundamentals and hands-on TypeScript experience.
- Experience shipping LLM-powered systems to real customers and operating them after launch.
- Deep hands-on experience with agent orchestration, tool calling, and context management.
- Experience with sandboxed code execution, inference infrastructure, or similarly demanding production systems.
- Ability to evaluate system quality beyond demos, investigate failures, and make sound technical decisions.
- Python experience and familiarity with evaluation frameworks are useful.
- Procurement experience is a plus.
Benefits
- Competitive salary and meaningful equity are offered.
- This is an on-site role in Berlin, Germany.
- Visa sponsorship is not available.