Software Engineer Robotics Research Tooling & Reliability
Role details
Job location
Tech stack
Job description
The AI Research Division of Agile Robots is looking for a Software Engineer (m/f/d) Robotics Research Tooling and Reliability who would take responsibility for the usability and reliability of the toolchain that our teams depend on daily, from dataset viewers, training dashboards, benchmark interfaces until the end-to-end test infrastructure that keeps them stable. The job is to make the tooling ecosystem trustworthy under pressure., * Tool Surfaces: Build and maintain UI, API, and CLI interfaces (dataset viewers, training dashboards, benchmark viewers) that researchers use for daily inspection and monitoring work.
- End-to-End Testing: Design and maintain E2E test suites that simulate real user workflows from data ingestion through training to inference, catching regressions before they reach production.
- Stress Testing: Run stress tests under realistic lab constraints (messy data, constrained compute, edge-case queries) to ensure tooling degrades gracefully rather than failing silently.
- Golden Path Maintenance: Maintain a reference test suite and nightly regression runs that define the canonical data-to-training-to-inference workflow and prevent drift.
- Demo Reliability: Own reliability checks and regression prevention for critical demos and evaluations, ensuring the toolchain survives the pressure of time-sensitive events., * Frontend Development: Working experience with a modern frontend framework for building data-intensive interfaces (dataset browsers, real-time dashboards, or annotation tools).
- Documentation Habits: Practice of maintaining clear, current documentation as a first-class engineering output rather than an afterthought.
Requirements
Do you have experience in Front-end development?, * Backend Engineering: Experience building production-quality backend services that maintain reliability under real load, not scripts or single-use tools.
- Testing Discipline: Hands-on experience designing integration and E2E test suites, with a debugging instinct that works across the full stack.
- User Empathy: Track record of building internal tooling that technical users actually adopt, with evidence of learning from usage patterns.