Deployment Engineer

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.

This is a hands-on role for someone who's equally comfortable reading a Python traceback, crimping an ethernet cable, arguing about GPU thermals, and talking a customer's IT contact through firewall whitelisting, often in the same afternoon.

• →

  Act as the first-line software responder for live robot operations. Review operator reports, assess safety and production impact, provide clear guidance, and drive remote recovery across robot services, servers, networking, and configuration.

• →

  Own production change management and incident evidence. Coordinate downtime and deploy/rollback station, inference, and model releases; and route clean handoffs to partner teams.

• →

  Stand up and maintain deployment sites. Network bring-up (captive portals, firewall/MAC whitelisting, token rotation, Tailscale, VLANs/SSIDs, WiFi and TCP tuning), inference server provisioning, and end-to-end validation that a robot is collecting/inferring correctly.

• →

  Own incident response. Carry the on-call pager, triage and lead incidents on incident.io, debug under pressure (motor/CAN bus timeouts, container crash-loops, upload/finalize-queue stalls, GPU-not-detected), and write the post-mortem and follow-ups.

• →

  Debug across the stack. Linux services and Docker container health, GPU/inference server issues, networking, and robot hardware (arm/motor swaps, cabling, sensors).

• →

  Be the on-site face for customers. Coordinate directly with customer IT and site contacts, schedule and run deployments, and represent Dyna professionally at sites, showrooms, and conferences.

• →

  Make the function scalable. Turn tribal knowledge into runbooks, networking/provisioning scripts, monitoring/dashboards, and onboarding so that no single site (or person) is a point of failure.

• →

  Close the loop with engineering. File clear, reproducible reports back to hardware/software/research teams and push for fleet-wide fixes (not just one-off patches).

• Strong Linux fundamentals: shell, systemd/services, SSH, log spelunking, and Docker.

• Solid networking skills: DHCP/DNS, firewalls, VPN (e.g. Tailscale/WireGuard), VLANs, WiFi behavior, and the patience to deal with hostile customer networks (captive portals, locked-down firewalls).

• Comfort with GPU/server hardware and inference workloads: drivers, nvidia-smi, thermal/power constraints, basic benchmarking.

• Hands-on hardware debugging instinct — willing to physically swap an arm, reseat a connector, or trace a flaky cable.

• Calm, structured incident response and clear written communication (status updates, post-mortems).

• Customer-facing maturity and willingness to travel to sites on short notice.

• Robotics or embedded experience: CAN bus / socketcan, motor controllers, sensors/cameras (e.g. GStreamer pipelines).

• Experience building observability (Grafana/Datadog) and writing deployment/automation scripts.

• Prior field/deployment, SRE, or solutions-engineering role at a hardware or robotics startup.

• Experience setting up on-call processes and runbooks from scratch.

• Location: in Redwood City, CA (Onsite 5 days/week + occasional travel)