The Facility Controls TeamThe Facility Controls team builds the control systems behind Fluidstack's data centers: real-time load control, MEP and behind-the-meter integration, automated commissioning, and autonomous deployment.
Examples of key problems the team is working on- Deliver the controls behind gigawatt-scale data centers this decade.
- Own real-time load control and MEP and behind-the-meter integration.
- Automate Level 4 and Level 5 commissioning.
- Drive autonomous, robotic deployment.
Role Scope- Build and own the facility side of the energy control contract. The production services that carry it: the protocol layer, the gateway service, and the API third-party plant controllers poll against.
- Design and evolve the messaging and state layer. It computes the facility's load intent from live telemetry and publishes it every second with a forecast ahead of it.
- Own the data contracts and engineering standards other teams build against. Keep them from drifting as the platform reaches new sites and new counterparties.
- Build the constraint execution path. It answers an incoming request with what the facility can actually achieve and by when, inside deadlines measured in seconds.
- Build the observability that proves conformance in production. Per-stage latency, staleness and watchdog health, so a missed budget is visible before it becomes an incident.
What We're Looking ForThe below is a starting point. We always make space for exceptional people, so if you don't fit this role exactly, tell us where you would.- You have worked with high-throughput messaging systems (NATS, Kafka, or equivalent) at real scale and you know where the failure modes are before they surface in production.
- You have designed time-series data models in ClickHouse, TimescaleDB, or a comparable system and you understand the tradeoffs between write throughput, query performance, and schema evolution.
- You have instrumented production services with observability tooling (Prometheus, Grafana, or equivalent) and you treat metrics and alerting as part of shipping, not something you add after the fact.
- You have built and owned production services that other teams depend on, and you do not ship code you would not want to be paged for at 3am.
- You design before you build. You can name the patterns you reach for and why, explain the alternatives you rejected, and point to a system you deliberately refactored because the original design stopped fitting the problem.
- You draw boundaries on purpose: interfaces, modules and failure domains, so a dependency going down degrades your service instead of taking it with it. The tests you write are the ones that catch the failures you actually fear.
- You have worked across the full signal chain from device to database, and you treat data integrity as non-negotiable. You work backwards from device-level constraints to build pipelines that are correct by design rather than merely functional, and you handle a stale value differently from a wrong one because both have burned you.
- You move toward a broken pipeline in production the same way you would move toward any other hard problem: with urgency and without drama.
- Bonus: Our stack (Go, NATS, Redis). Industrial and utility protocols (DNP3, Modbus TCP, OPC UA). Kubernetes and ArgoCD for production service deployment. Real-time control and dispatch systems (SCADA, EMS, power plant controllers). Real-time dashboarding (Grafana or equivalent).
We are committed to pay equity and transparency.
You will receive a confirmation email once your application has successfully been accepted. If there is an error with your submission and you
did not receive a confirmation email, please email [redacted] with your resume/CV, the role you've applied for, and the date you submitted your application-- someone from our recruiting team will be in touch.
