> Markdown version of [/jobs/ext/4774491-mts-infrastructure-engineer-field](https://www.wearedevelopers.com/jobs/ext/4774491-mts-infrastructure-engineer-field). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # MTS - Infrastructure Engineer, Field - **Company:** Ironsite AI Inc. - **Location:** San Francisco, CA, United States - **Experience:** Experienced - **Salary:** $225,000 - $275,000 - **Contract:** Permanent contract - **Skills:** OpenTofu, Artificial Intelligence, C++ (Programming Language), Data Compression, Computer Programming, Software Debugging, Linux, Failover, FFmpeg, Firmware, Hardware Interface Design, Virtual Private Networks (VPN), Mesh Networking, Network Monitoring, Ansible, Prometheus, Remote Infrastructure Management, Simple Network Management Protocols, Video Encoding, Wide Area Networks, Wi-Fi Technology, Network Routers, Computer Network Operations, Grafana, Terraform - **Published:** November 18, 2026 - **Apply:** https://www.careerbuilder.com/job-details/mts-infrastructure-engineer-field-san-francisco-ca--9c421ac0-0997-4ef3-b700-f03044189ebf ## About the Role Data Modeling, Sports, Network Operations Center, Construction, Philosophy, Hospital, Artificial Intelligence (AI), Liquified Natural Gas (LNG), Construction Projects, Operating Systems, * Strong programming in Rust, C++, or C, with production software on resource-constrained Linux (video pipelines, networking, long-running services). * Hands-on experience designing and running networks in difficult environments, including WAN/LAN, Wi-Fi planning, satellite and cellular backhaul, VPNs and overlay networks, and failover. * Deep Linux and systems skills: building OS images, debugging kernel and driver issues, and working within storage, power, and thermal limits on edge hardware. * Infrastructure as code (Nix, Terraform/OpenTofu, Ansible, or similar); every box rebuildable from the repo. * Experience running device fleets with staged rollouts, remote management, monitoring, and alerting. * Willing to travel to jobsites to bring up networks and debug in person. * Typically 4+ years in infrastructure, networking, or fleet operations; CS, EE, or related degree, or equivalent experience., * NixOS or other reproducible, declarative systems in production. * IoT, industrial, or OT networking: PoE, ruggedized gear, and harsh RF environments. * High-throughput transfer over unreliable links: store-and-forward, resumable and delta sync. * Video encoding and processing: hardware encoders, FFmpeg or GStreamer pipelines. * Hardware interfaces (GPIO, ADC, PWM, serial and fieldbus) and working closely with EE and firmware teams. * Network monitoring and telemetry across distributed sites: SNMP or streaming telemetry, flow data, synthetic probes, and Prometheus-compatible metrics with Grafana. * Security fundamentals for edge fleets: device identity, certificate rotation, and secure remote access. * Field deployments in construction, energy, telecom, or similar industries. ## Description We're hiring an engineer to own field infrastructure: the networks and edge compute that keep thousands of wearable cameras recording, charging, and uploading on some of the largest jobsites in the country. Power is temporary, connectivity is satellite or cellular, and the site looks different every week. Every shift's footage needs to reach the cloud before the next morning's report. You'll own the path from the device to our cloud, starting with the network and compute in the on-site rack we're building, a ruggedized telecom-style unit with edge compute, industrial switching, satellite backhaul, power, battery and charging drawers, and I/O to site equipment. You'll write the software that runs on it, ship the updates that reach it, and build the network monitoring that catches problems early., * Edge workloads. Video compression and pre-processing, upload scheduling, and the services that move a site's footage, written in Rust, C++, and C to run on tight power and bandwidth. * The on-site rack. The compute and network side of our rack, built with the hardware and firmware teams. Industrial switching, backhaul, power and battery telemetry, and digital, analog, and PWM I/O to peripherals. * Edge fleet. Reproducible NixOS images, staged rollouts with canaries and rollback, and auto-update that works when a site has been offline for days. * Device-to-cloud ingestion. A full shift of video uploaded overnight over constrained, metered links, with resumable transfers, bandwidth shaping, and on-site buffering so nothing gets dropped. * Site launches. A repeatable playbook for bringing up a new jobsite, from hardware kits and charging to network bring-up and remote checks before the first shift. * Network observability. Per-site uplink latency, loss, and throughput, backhaul failovers, Wi-Fi and RF health, switch and port telemetry, and upload rates, with alerts within minutes when a link drops or saturates. You'll also work with hardware and platform on fleet-wide device monitoring. * Inventory and lifecycle. Every camera, router, and edge box tracked from shipping to decommission, with firmware, OS, kernels, and certificates kept current., * Offline is the default. Sites drop connectivity for hours or days at a time. Updates, telemetry, and uploads have to queue, resume, and catch up on their own. * Sites change every week. Steel, concrete, and new floors break RF plans as the building goes up, so networks and tooling have to adapt with the site. * No IT on site. Nobody on site can debug a network for us. If it can't be fixed remotely, it's a truck roll and a lost day of footage. * A fixed power budget. The rack runs on site power and batteries, so compression and pre-processing have to fit a fixed power and thermal envelope and still keep up with the fleet. * Fleet scale. Thousands of devices across dozens of sites in seven-plus states, all reproducible, patched, and accounted for.