Infrastructure Engineer (Member of Technical Staff)

Ironsite AI•San Francisco, CA

138d•Onsite

About The Position

Ironsite is building the intelligence layer for the physical world by designing wearable hardware, deploying it alongside craft workers, and transforming shift footage into next-morning reports. Our internal team and purpose-built models label data overnight to deliver actionable insights to superintendents by 5 AM. We are accelerating the speed, efficiency, and predictability of construction for complex, mission-critical infrastructure projects by training AI models on egocentric construction footage and labor productivity data. Our philosophy is pro-worker, believing technology should empower, not replace, the workforce. We aim to give craft workers and project leaders better on-site visibility and create a system where the reality of construction is available to project leaders. Ironsite is deployed on major construction projects, having captured over 100,000 hours of footage across seven states and processing thousands of hours daily, with a worker opt-out rate below two percent. Our workforce-first architecture anonymizes devices, captures no audio, and never releases raw video. We are backed by leading investors and operators in technology and construction. Long-term, we see Ironsite as the foundation for the future of construction, acting as the operating system for how the physical world is built, fostering respect for the workforce, and enabling autonomous construction.

Requirements

Deep infrastructure experience across cloud, distributed systems, and ideally, the intersection of the two with hardware and edge devices.
Experience making judgment calls on when to use managed services vs. build in-house, optimize for cost vs. latency, and invest in abstraction vs. ship simple solutions.
Track record of architecting and shipping data or ML training pipelines that scaled by orders of magnitude in throughput, storage, or user count, without over-engineering or accumulating dead architectural weight.
Strong systems thinking across the full stack, from IoT/embedded to cloud/HPC, with the judgment to understand how choices in one layer propagate.
Comfort operating at the edge of what's known, solving infrastructure problems where the right answer isn't readily available.
Typically 8+ years of experience, with at least 3-4 spent operating at Staff/Principal level.

Nice To Haves

Deep experience with edge ML inference and neural accelerator chips.
Background across IoT networking and traditional multi-environment HPC networking.
A history of owning the full lifecycle of systems across hardware, firmware, and software integration.
Strong instincts for observability across disciplines.
Experience with security patterns for edge devices and video data across long lifecycles.
Genuine excitement about industries that have been underserved by software, and interest in building the infrastructure for the physical world.

Responsibilities

Build the systems that turn Ironsite's data into the intelligence layer.
Own the backbone connecting wearable devices to AI systems.
Develop the device-to-cloud ingestion layer for reliable data transfer from ruggedized wearable devices in harsh environments.
Create petabyte-scale video pipelines for ingesting, processing, storing, and serving millions of hours of video.
Deploy and optimize vision-language model (VLM) inference on neural accelerator chips across edge devices, on-premises rigs, and cloud clusters.
Build the research compute platform and multi-environment compute orchestrator for AI team training and evaluation.
Develop automated reporting infrastructure for delivering AI-powered productivity insights.
Implement privacy and compliance controls, including automated face and phone blurring and worker anonymization.
Ensure edge reliability at scale through monitoring, observability, and resilience patterns for the device fleet.