Plant Controls Lead (EMS/PPC)

Fluidstack•Austin, TX

10d•$250,000 - $330,000•Remote

About The Position

Fluidstack is building civilization-scale infrastructure for AI and is singularly focused on delivering 10 to 100s of GWs of compute faster than anyone else. This role is part of the Power Generation Team, which is working on key problems such as building an islanded grid at gigawatt scale, getting power online in months, creating repeatable power blocks, and designing a grid where the largest load answers back. The team operates with a philosophy of full autonomy, insane urgency, reasoning from first principles, a love of the game, and a focus on building something that matters. The Plant Controls Lead will own the control stack for fully off-grid campuses, including the site energy management system, the plant controller for each power block, and the dispatch logic across PV, BESS, and gas. This role involves turning control philosophy into sequences of operation, managing the control contract between power plants and data centers, deciding on component procurement versus in-house development, building controls test capabilities, leading field commissioning and tuning, and hiring controls software and test engineers.

Requirements

Owned the plant controller or EMS on a utility-scale hybrid plant combining at least two of PV, BESS and thermal generation, and can name the sites and the megawatts it dispatches today.
Written sequences of operation for source transitions, blackstart or islanding, and have stood on site when they ran against energized equipment, including the times they did not work.
Shipped control software as a product, version-controlled, tested and released to running plants, and know what changes when the loop closes on a real turbine or inverter fleet.
Brought a third-party controller or component into a live plant to hit a date, held the vendor to a specification through factory and site acceptance, and later replaced it with your own code without an outage.
Designed an interface between two control systems owned by different parties, with the data, timing and failure behavior written down and tested, not left to whoever integrated it.
Tuned a plant in the field, frequency and voltage response, ramp rates and load steps, and can explain the tradeoff behind each setting left in place.
Write clearly enough that an OEM controls engineer, a commissioning agent and a CEO each get the decision they need from the same document.

Nice To Haves

Islanded or weak-grid operation (island utilities, mining and remote microgrids, offshore).
Real-time simulation and hardware-in-the-loop (RTDS, OPAL-RT, Typhoon HIL).
Commercial plant controller platforms (FlexGen HybridOS, Tesla, Schneider EcoStruxure, SEL RTAC).
Industrial control protocols and OT security (DNP3, IEC 61850, IEC 62443).
Real-time and control software languages (Go, Rust or C++, IEC 61131-3).

Responsibilities

Own the control stack for fully off-grid campuses, the site energy management system, the plant controller for each power block and the dispatch logic across PV, BESS and gas, as one standard architecture that scales from a single block to a GW-scale campus energized in six months.
Turn the control philosophy into sequences of operation our own code executes, covering every operating mode, source transition, blackstart path, emergency shed and degraded state, with each one proven in simulation and hardware-in-the-loop before the plant is energized.
Hold the generation side of the control contract between the power plant and the data center, and extend it from one plant and one facility to many blocks feeding many buildings, with load reduction deadlines measured in seconds that both sides are measured against in production.
Decide which components we bring in to energize the first campuses faster and which we build first, then replace each bought piece with our own code on a planned schedule, holding every supplier to our specification and conformance tests so Fluidstack keeps its configuration, its data and a clean exit.
Build the controls test capability, a plant model library and hardware-in-the-loop bench where any controller, vendor or our own, is tested against our plant before it ships to site.
Lead controls commissioning and tuning in the field, through frequency and voltage response, load-step testing against real AI load profiles and every mode transition, and fold each finding back into the standard so the next campus starts tuned.
Hire the controls software and test engineers who build the stack with you, sequenced against the components we replace first.