Senior Software Engineer, Fleet Connectivity
Owns network performance and connectivity for Dyna Robotics' robot fleet: building monitoring, automated failover, and self-healing diagnostics, tuning low-latency teleoperation links, and handling customer site bring-up. Core tech: Linux networking, TCP/UDP, Wireshark/tcpdump, WireGuard VPNs, and production code in C++, Rust, or Python.
- Compensation
- Not disclosed
- Location
- Redwood City, CA
- Work setup
- Onsite
- Seniority
- Senior
- Employment
- Not specified
- Experience
- Not specified
- Category
- Software engineering
- Posted
Skills
Must have
- Hands-on experience debugging real-world networks: WiFi/RF troubleshooting and packet-level diagnostics (Wireshark, tcpdump, iperf)
- Deep understanding of TCP, UDP, and how they behave on lossy, high-jitter links
- Strong Linux networking skills: routing, traffic shaping with tc, firewalling, VPNs such as WireGuard
- Track record of shipping production-quality software in C++, Rust, or Python
- Comfort working in the field and communicating clearly under pressure
- Degree in CS, EE, or a related field, or equivalent practical experience
Nice to have
- Experience with real-time media or low-latency transport: WebRTC, RTP, adaptive bitrate, congestion control, jitter buffers, or UDP-based protocols like QUIC or SRT
- Experience managing connectivity for a fleet of edge devices (robots, vehicles, IoT) deployed on networks you don't own
- Hands-on work with cellular/LTE/5G failover hardware, link bonding, or multipath networking
- Experience with pub/sub or messaging middleware (ZeroMQ, ROS 2, DDS, or similar), especially over WiFi or WAN links
- Experience with network hardware: managed switches, access points, mesh networking, cellular routers
- Experience supporting live events, demos, or on-site customer deployments where network reliability was your responsibility
Full description
Dyna Robotics builds general-purpose robots powered by a proprietary embodied AI foundation model with top-in-industry generalization and real-world performance. Already deployed with customers across multiple industries, our robots do commercial-grade work in the physical world. Our team comes from Google DeepMind, Meta, and Cruise, and we're backed by CRV, First Round, and other leading investors.
The Role
Dyna's robots only create value when they're reliably connected. Robotics isn't like typical enterprise software: latency that an ordinary business app would shrug off turns teleoperation and remote supervision into jerky, unsafe, or outright failed robot behavior. A degraded link here isn't a background IT ticket: it can be the difference between a robot finishing a task and stalling mid-motion. This role owns the technical side of network performance in two settings: making sure a demo never drops mid-pitch, and making sure a deployed robot stays online, monitorable, and controllable on network infrastructure Dyna doesn't own or control. This is a build role, not a network-admin or support role: you'll design and build the systems (monitoring, automated failover, self-healing diagnostics) that make connectivity reliable by default, so stability is a property of the system rather than something a person has to catch and fix after it breaks.
What You'll Do
- Own network performance for remote teleoperation. Measure and minimize end-to-end latency and jitter, tune how video and control streams are transported and prioritized, and make sure sessions degrade gracefully (lower bitrate, prioritized control traffic, safe fallback behavior) instead of failing when the link gets worse. A reliable, low-latency link is what keeps a human safely in the loop.
- Build the fleet connectivity stack. Monitoring, diagnostics, and automated failover software that keeps every robot's connection healthy in production, turning a dropped connection into an automatic root cause and recovery, not a 2am page to go run tcpdump on-site.
- Own the network side of site bring-up. Build tooling that assesses a customer's network before a robot ever arrives, then work with the deployment team to configure the right failover, QoS, and VPN setup for that site.
- Design connectivity for real-world chaos: warehouse WiFi dead zones, cellular failover, and VPN-gated corporate networks with firewall rules you can't change.
- Diagnose field issues on networks you don't control, then build the fix so the same failure never happens twice.
- Make the office and demo network rock-solid by default, so a live demo never drops in front of a customer or investor.