Summary
✨ AI‑Generated
A founding engineering role focused on building the backend orchestration layer for a distributed GPU compute platform. You will design workload placement and matching logic, integrate external compute providers through APIs and networking, manage scaling and operations, and ensure workloads start, run reliably, and terminate correctly. This is a highly hands-on role with close user interaction and broad technical ownership.
Highlights
Founding-level ownership, direct influence over orchestration architecture, hands-on work with distributed compute infrastructure, responsibility for workload placement and operations, and the opportunity to build core systems from an early stage.
Description
Founding Engineer, OrchestrationBuild the orchestration backend of Pyrois: the matchmaker that places every workload, the connections to our providers' data centres, and the operations that keep customer workloads running.
Go first, hands-on, close to users.
About PyroisPyrois is one interface to the world's GPUs.
We aggregate compute across neoclouds and data centres and route every workload to its best price and match, continuously.
Customers stay sovereign, operate at best-market rates and source the compute they need, when they need it.
We're a VC-backed, pre-seed company, hiring our first ten people now.
What you'll doBuild the matchmaker.
Scoring, constraints, the audit trail behind every placement, and the explanation when nothing fits.Connect data centres.
Bring new providers and their capacity onto the platform, from their APIs to the networking.Keep customer workloads running.
Own orchestration, scaling and operations, so a placed workload starts, stays up and stops cleanly.Extend the MCP surface so agents can provision, quote and manage their own compute.Start the telemetry loop.
Collect what each placed workload actually needed, so the next placement is cheaper.Later: MLOps integrations (Ray, Kubeflow, MLflow, Metaflow).What you bringYou've built and run orchestration or scheduling systems in production.You're strong in Go and comfortable in other languages when the job needs them.You've operated infrastructure at scale: capacity, incidents, on-call.You're hands-on, and you talk to users and stakeholders directly.RequirementsMust-have
Built and operated orchestration or scheduling systems in production (for example Kubernetes, Nomad or an in-house scheduler)Production experience writing GoOperated infrastructure at scale: capacity planning, incident response, on-callWorked directly with users, customers or stakeholders on technical problemsNice-to-have
Built placement, matching or scheduling logic (bin-packing, constraints, auctions)Integrated cloud, bare-metal or data centre providers through their APIs, including networkingExperience with ML or inference workloads in operationsWorks in other languages besides Go, such as Python or TypeScriptCompensationGBP 75'000 - 95'000, London
BenefitsGenerous early-stage equity through our employee stock option pool25 days of paid time offRegular team off-sitesThe AI tools you work with, paid forYour choice of laptop and equipmentFlexible working hoursHow we workSome roles are hired on-site in Zürich, where we work together three days a week in the office.
Other roles are fully remote and we meet regularly with the whole team (travel paid).
We have four hours of overlap with the Zürich office every day, to be able to collaborate.
Where you can work fromWe employ through Remote.com as our employer of record.
Wherever you live, you get a local employment contract, payroll and benefits under your country's law, without us opening a company there.
