FluidStack

Software Engineer, Facility Software Automation

Austin, TX$200k–$250kfulltimemidAdded today

About this role

Fluidstack seeks a Software Engineer to build and own production telemetry and automation services that power AI compute infrastructure at massive scale. You'll design data pipelines, messaging systems, and APIs that collect and act on real-time signals across a rapidly expanding global fleet of facilities.

What you'll do

Build and maintain production telemetry collection, ingestion, and automation services that downstream teams depend on
Design and evolve messaging and time-series data layers for real-time device signal ingestion across the compute fleet
Own software infrastructure and data lake, establishing data contracts and engineering standards for other teams
Create APIs and dashboards surfacing telemetry to operations, commissioning, and leadership teams
Drive integration between device-level data streams and automation APIs as the platform scales to new sites
Ensure telemetry pipeline health and data integrity at gigawatt scale

What they're looking for

Production service ownership and deployment
High-throughput messaging systems (NATS, Kafka, or equivalent)
Time-series database design (ClickHouse, TimescaleDB, or similar)
Observability and monitoring (Prometheus, Grafana)
Full signal chain architecture from device to database
Data pipeline design and optimization
API design and real-time data systems
Software infrastructure and data engineering

Benefits

Work on civilization-scale AI infrastructure
Extreme ownership with full autonomy over end-to-end systems
High-velocity environment with first-principles problem solving
Impact on frontier AI deployment and human freedom
Opportunity to work with cutting-edge compute infrastructure

Likely interview questions

Describe a time you owned a production service that other teams depended on—what made you confident shipping it, and how did you handle incidents?
Walk us through a high-throughput messaging system you've operated at scale: what failure modes surprised you, and how did you design around them?

How would you design a time-series schema to handle device signals from thousands of facilities with varying collection rates, and what tradeoffs would you make between write throughput and query performance?
Tell us about a telemetry or data pipeline you've built—how did you ensure data integrity when the system scaled, and what observability did you bake in from day one?
Describe how you'd approach designing a telemetry platform backwards from device-level constraints: what would you validate before writing code?
What's an example of taking on scope outside your core role to unblock a team or accelerate a project, and how did you prioritize it?
How do you think about the relationship between signal quality and automation reliability in a system where missed alerts could impact customer production workloads?

Unlock all 7 questions free — and practice them live →

Apply on the employer's site →

Opens the official application on the employer’s site. No login required.