Staff Robotics Software Engineer, Teleoperations Architecture
Architects real-time teleoperation systems for humanoid robots (e.g., Agility’s Digit) to enable human-guided manipulation tasks in warehouses/manufacturing, focusing on low-latency C++/Rust software for hardware-AI integration, data pipelines, and XR interfaces.
- Compensation
- $190K–$296K / yr
- Location
- Hybrid- Any Office (Fremont, CA, Salem, OR, or Pittsburgh, PA)
- Work setup
- Hybrid
- Seniority
- Staff
- Employment
- Full time
- Experience
- 7+ years
- Category
- Software engineering
- Posted
Skills
Nice to have
- Domain Knowledge: Familiarity with whole-body kinematics, motion retargeting, rigid body dynamics, or real-time control systems.
- Human-Machine Interfaces: Hands-on experience integrating human input systems such as Virtual Reality (VR), motion capture, wearable devices, or remote control interfaces.
- XR Frameworks: Experience building applications with real-time rendering or XR frameworks (e.g., Unity, Unreal, Godot, SteamVR, OpenXR).
- 3D Visualization: Experience with advanced 3D visualization techniques such as neural radiance fields, Gaussian splatting, or other scene reconstruction methods.
- Data Pipelines: Familiarity with large-scale data collection pipelines for learning from demonstration.
Full description
Agility’s commercially deployed humanoids operate alongside teams in warehouses, manufacturing facilities, and distribution centers—tackling physically demanding and repetitive tasks while enabling workers to focus on higher-value work. With industry-leading safety standards and years of proven deployment data, we're pioneering a new era of automation that enhances human potential.
About The Role
You will be the architectural force behind the systems that teach Digit how to interact with the physical world. As a Staff Robotics Software Engineer on the AI team, you will architect, build, and scale the production-quality software foundation that enables Digit to execute complex, whole-body manipulation tasks in real-world environments. You will work at the critical intersection of AI, controls, hardware, and operations to create a seamless, low-latency platform for human demonstration data collection. Your software architecture will directly unlock our ability to scale the world's first production humanoid robot fleet and revolutionize the global labor market.
About The Work
- Architect from the Ground Up: Own the end-to-end software architecture of our data collection and teleoperation systems, ensuring highly modular, scalable, and predictable behavior with clear system boundaries.
- Optimize for Performance: Build the connective tissue for real-time robot execution by integrating and optimizing real-time control, streaming, and communication paths to meet strict latency and bandwidth targets.
- Design for the Future: Anticipate upcoming robot hardware, sensing, and end-effector changes. Design resilient software systems that can seamlessly adapt to new interfaces and capabilities without requiring foundational architectural rewrites.
- Drive Validation & Reliability: Develop comprehensive validation strategies across simulation and physical hardware, including performance benchmarks, fault handling, and regression testing.
- Data-Driven Iteration: Analyze system telemetry and demonstration data to continuously improve operator efficiency, system responsiveness, and overall task success rates.
About You
- 7+ years of experience in software engineering for robotics, real-time systems, or highly complex hardware-software integrations.
- Exceptional software architecture skills, with a proven ability to design and build scalable, low-latency systems from scratch without foundational missteps.
- Demonstrated track record of operating as a highly autonomous, independent executor capable of delivering reliable, production-quality software.
- Experience designing low-latency communication and streaming systems for real-time applications.
- Strong systems intuition with the ability to trace and debug performance issues across complex software, networking, and hardware boundaries.