Jobs at Fluidstack

Own roadmap and product direction for ontology and data products. Translate messy operational domains into data models and specs. Ship end-to-end with engineers, validate with stakeholders, and define metrics proving operational leverage.

Lead end-to-end R&D program plans for hardware and infrastructure: scope, sequencing, owners, and critical path from concept through prototype, qualification, and production handoff. Run NPI programs, coordinate design reviews, procurement, lab/test capacity, surface risks, drive cross-functional execution with manufacturing, supply chain, and deployment, and build leadership reporting while optimizing for velocity.

Own and scale the GPU compute fleet: build metrics, alerting, observability, and repair automation; design GPU qualification/burn-in pipelines; own Redfish/BMC tooling and firmware telemetry; run incidents, eliminate toil, and deliver end-to-end reliability and orchestration for Kubernetes and bare-metal compute at hyperscale.

Build and operate large-scale GPU compute fleet tooling: metrics pipelines, alerting, Redfish/BMC telemetry, repair/RMA automation, GPU qualification, and orchestration for Kubernetes and bare-metal deployments. Own reliability, scalability, incident response, and automation to minimize manual toil.

Model and track fleet capacity (power, space, cooling, compute), forecast constraints, connect deployment schedules to sellable capacity, run scenario analyses to inform allocation and deal decisions, and build executive-ready reporting.

Build AI-powered internal tools and shared AI infrastructure (agents and skills), integrate LLMs with internal systems, design and ship web applications, and enable cross-functional adoption through documentation and support.

Build and operate GPU fleet infrastructure: metrics, alerting, Redfish/BMC tooling, repair and qualification pipelines, automation for failure detection through RMA, and ensure reliability, scalability, and observability for Kubernetes-orchestrated and bare-metal GPU deployments.

Lead technical execution of large-scale network deployments: define ZTP and validation playbooks, set acceptance criteria, resolve escalations across links and fabrics, and automate turn-up processes to eliminate manual repetition.

Lead product direction for ontology and data products that model and support compute production. Translate messy operational domains into data models, specs, and product surfaces; drive technical tradeoffs and ship end-to-end; run stakeholder discovery and validate impact; define and track success metrics to optimize product outcomes.

Design and develop the mechanical aspects of data-center robotics including end effectors, docking, and robot-plant interfaces. Rapidly prototype, test to failure, instrument trials, and productionize hardware with DFM, reliability, and spare strategies while collaborating closely with controls and software teams.

Own the steel category for structural and fabricated assemblies: source regional fabrication capacity, qualify fabricators for quality and throughput, negotiate raw material and fabrication pricing resilient to commodity swings, balance weekly loads across shops, and manage commodity exposure through indexing or hedging.

Serve as Fluidstack's public face at town halls, hearings, and community events; build and maintain relationships with officials, utilities, nonprofits, and stakeholders; identify and mitigate political and reputational risk; deploy community benefit programs and local hiring/education partnerships; manage external lobbyists, consultants, and PR agencies across concurrent site campaigns.

Manage deployment projects for controls and sensor systems across multiple data center sites: scope, schedule, vendors, installations, integration with construction and commissioning, vendor relationships, and punch-list closure to completion.

Build and maintain demand and lead-time forecasts across critical equipment categories, track supplier capacity and delivery risk, model cost/timing trade-offs for large equipment commitments, reconcile supplier/shipment/inventory data into actionable weekly signals, and create forecasting tooling and dashboards to surface disruption ahead of schedule.

Design and implement demand management control logic that flexes large IT and facility loads against grid signals and prices, integrate with utility and market signals, validate strategies in simulation, and deploy settlement-grade, fleet-wide control strategies without impacting SLAs.

Engineer a virtualized PLC platform: port and validate control applications to software runtimes while preserving determinism and safety; build test harnesses to prove virtual controls against hardware behavior; define reference architecture for networking, redundancy, and failover; document control narratives for engineers and operators.

Own supplier integration into production: onboarding, data connections, launch readiness, performance management, scorecards, escalations, and recovery plans. Ensure forecasts, releases, and ASNs flow into ERP/MRP without spreadsheets and coordinate cross-functional supplier launches.

Partner with leaders to define pay strategy, own annual and off-cycle comp planning, conduct market benchmarking, design compensation and incentive programs, build compensation tools/models/dashboards, support performance and offer decisions, and translate data into actionable recommendations to retain and reward talent across the company.

Own compute turn-up from facility availability to ready-for-service: qualify racks at scale (BMC/BIOS/firmware, burn-in, node and cluster validation), automate qualification and provisioning workflows, triage hardware failures and drive RMAs, run mostly-remote turn-up with brief on-site pulses, and partner across network, ICT, ops, and hardware teams.

Design, program, and configure BMS controls for mechanical and electrical plant. Build sequences, point lists, and graphics; integrate chillers, CDUs, power monitoring, and third-party equipment over BACnet/Modbus. Commission systems, diagnose failures, and version and document logic for maintainability.