Embedded Linux Engineer — On-Site AI Workstation & Jetson Setup Engineer — Dubai (Short-Term Contract)

Canopy Artificial Intelligence is deploying computer vision infrastructure for construction site monitoring. We need someone in Dubai to physically commission two machines and hand them back to us as remotely accessible, GPU-ready boxes.Scope of workUnbox, assemble, and bench-test an RTX PRO 6000 workstation (PSU, cooling, and thermal validation under sustained GPU load)Install Ubuntu Server 24.04, NVIDIA driver stack, CUDA toolkit, Docker, and NVIDIA Container ToolkitFlash an NVIDIA Jetson AGX Thor with the current JetPack release via SDK ManagerNetwork configuration: static IPs or DHCP reservations, port forwarding or a mesh VPN overlay (Tailscale/WireGuard), verified stable connectivityConfigure hardened SSH access for our team: key-only auth, no password login, no root login, fail2ban or equivalentValidate and hand over: nvidia-smi and jtop reporting correctly, a sample CUDA and TensorRT workload running end-to-end, unattended reboot survivalRequirementsPhysically located in Dubai and able to work on site at Palm Jabal AliDemonstrable experience flashing Jetson devices (Orin, Thor, or Xavier)Comfortable with Linux server administration, systemd, and networking fundamentalsPrior work with NVIDIA driver/CUDA installs on workstation-class GPUsNice to haveTensorRT, GStreamer, or DeepStream exposureIP camera and RTSP/SRT streaming experienceRack, UPS, or edge-deployment backgroundEngagement2 weeks of work. Remote follow-up support for 1 week after handover is expected as part of the fee.