Safety Operations Lead

Who we are Atoms is building the machines that power the next era of progress. Over the last decade, software has transformed the digital world. But the physical world, where food is made, minerals are mined, goods are moved, and industries are run, remains far less intelligent, far less efficient, and far more constrained. We're changing that. Atoms builds Physical AI- real-world robots for the industries that move civilization forward, starting with food, mining, and transport. Our systems are designed to understand, predict, and control the real world with precision, turning complex physical operations into something more reliable, more scalable, and more productive. This work requires more than robotics. It requires deep integration across hardware, software, AI, operations, manufacturing, and real estate. We don't just build machines in a lab. We deploy them into real environments, operate them, learn from them, and improve them until they work at scale. We are roboticists, engineers, operators, and builders. We believe the next great technology companies will not only transform information, but the physical systems that shape everyday life. If you want to work on hard problems with real-world impact, join us. About the role Our safety practice is not a document exercise - it is an operating discipline that has to run at the speed of the fleet. This role owns that discipline. You will stand up and run the operational safety machinery: how we manage incidents, how we train and qualify the people in and around our vehicles, how field events get triaged and routed, and how investigations turn into engineering change. Today these functions exist on a smaller scale. Your job is to make them a coherent system that scales as we add vehicles, routes, and customers. This is a hands-on role. You will write the next version of most of these processes yourself, run the first dozen investigations yourself, and only then hire against the parts that need scale. What you'll do Incident management. Own the end-to-end response process: detection, notification, severity classification, on-scene coordination, containment decisions, root cause investigation, corrective action, and closure. Recommend who should have the authority to hold operations and under what conditions, and make sure that authority is real. Event triage. Build the pipeline that turns raw field telemetry, disengagements, operator reports, and near-miss data into a ranked, actionable queue. Set the taxonomy and severity model. Own the triage rate, the backlog, and the time-to-disposition - and make those numbers visible to engineering and leadership. Investigations. Lead investigations of collisions, disengagements, and anomalous behavior. Drive them to a defensible causal account, not a narrative. Track corrective actions through to verified closure. Training and qualification. Design the training and recurrent-qualification program for safety drivers, autonomous vehicle operators, and remote support staff. Own the curriculum, the practical evaluation, and the qualification records. Safety data and reporting. Own the operational safety metrics set and the reporting cadence, and make sure our internal records can withstand outside scrutiny. Feedback into engineering. Be the channel that carries field reality into the product. Partner with the Head of Safety so that operational evidence feeds hazard analysis and change assessment rather than sitting in a separate system. What success looks like First 45 days: Incident response process documented, socialized, and exercised at least once in a tabletop. Event taxonomy and severity model in place with the current event backlog classified against it. A clear-eyed written assessment of where our operational safety practice is weakest. First 3 months: Triage running on a predictable cadence with published metrics. Investigation process producing written reports with tracked corrective actions. Training program rebuilt with documented qualification records for every person on duty. You have identified the AI-native tooling that will scale this work. First 12 months: The system runs without you personally in every loop. That tooling is implemented and in daily use. You know what the next hire needs to be and why. Working AI-native We expect this function to run leaner than it traditionally has run, and we expect AI tooling to be the reason. We are looking for someone who uses LLM-based tooling as part of the actual work - querying and summarizing large volumes of event data, drafting and maintaining investigation reports and training material, doing first-pass classification of an event backlog, pattern-finding across incidents that a human reading one at a time would miss - and who builds their own small tools rather than waiting on a roadmap. The boundary matters as much as the leverage: AI output is a hypothesis and a drafting accelerant, never the basis for a safety determination or a reg #J-18808-Ljbffr Atoms