3+ years of experience in building production-level real-time networked systems such as video streaming or remote desktop applications.
Hands-on expertise with low-latency media transport protocols like WebRTC or RTP/SRTP.
Proven ability to measure and reduce latency in production systems.
Experience in creating polished user interfaces for tools used extensively by operators.
Proficiency in at least one programming language: TypeScript, Go, Rust, C++, or Python.
Ability to deploy and debug software on remote Linux machines.
Responsibilities
Build real-time transport systems for low-latency video and control commands between robots and operators.
Manage end-to-end latency budgets, analyzing and optimizing every millisecond of delay.
Ensure robustness of connections through NAT traversal, congestion handling, and quality indicators.
Design and develop the operator interface, focusing on usability and efficiency for daily tasks.
Create operator-facing tools for managing robot sessions and fleet visibility.
Oversee on-site recording of synchronized streams and metadata, ensuring reliable uploads.
Establish infrastructure for data storage and access related to robot teleoperation episodes.
Benefits
Flexible work environment with opportunities for remote collaboration.
Access to cutting-edge technology and tools in robotics and teleoperation.
Support for professional development and continuous learning.
Collaborative team culture focused on innovation and problem-solving.
Full Job Description
The role
Our operators control robots across countries and oceans. You'll own the link between them: getting multiple camera streams to the operator and commands back to the robot with as little delay and jitter as physics allows, and building the interface operators use all day. Because every byte you stream is also training data, you'll own recording it on-site and getting it reliably off-site too.
What you'll do
Build the real-time transport between robot sites and operator stations: low-latency video (multiple cameras, possibly depth), robot state, and control commands over the public internet.
Own the end-to-end latency budget. Instrument glass-to-glass and command-to-motion latency, find where every millisecond goes, and cut it: encoding settings, transport choice, routing, jitter buffers, relay placement.
Make the link robust: NAT traversal, congestion and packet loss, bandwidth adaptation, reconnection, and clear signals to the operator and robot when quality degrades.
Design and build the operator interface: camera layouts, robot state and health, task instructions, latency and connection indicators, and fast workflows for starting, labeling, and ending episodes.
Build operator-facing tooling: queueing operators to robots across sites, session handoff, and basic fleet visibility.
Own on-site recording: capture synchronized streams and metadata into a well-defined episode format, buffer locally, and upload reliably over constrained site networks.
Set up the infrastructure for data storage and access of robot teleop episodes.
What we're looking for
3+ years building production level real-time networked systems: video streaming, video conferencing, cloud gaming, remote desktop, live broadcast, or similar.
Hands-on experience with low-latency media transport (e.g., WebRTC, RTP/SRTP, QUIC/UDP-based protocols) and video encoding trade-offs.
You've measured and reduced latency in a production system and can explain where the time went.
Comfortable building a polished interface yourself, web or native, with good product instincts for people who use a tool eight hours a day.
Strong in at least one of TypeScript, Go, Rust, C++, or Python; comfortable across client, server, and edge devices.
Comfortable deploying and debugging software on remote Linux machines you can't physically touch.
Nice to have
Experience with robotics, teleoperation, drones, or other remote-control systems.
Hardware video encode/decode on edge devices; multi-camera synchronization.
Built data pipelines that move large volumes of sensor or video data from the edge to cloud storage.
Experience with global network deployment: relays/TURN, choosing regions, measuring paths between countries.
Game development, VR/XR, or other latency-sensitive interaction design.
If you don't check every box above but believe you can do the job, we encourage you to apply - we're looking for capability and trajectory, not a perfect checklist match.
