Firmus Technologies
Firmus Technologies is a global leader pioneering the development and operation of efficient AI infrastructure across Asia Pacific.
Founded in Australia in 2019, our mission is to create the most efficient AI infrastructure by combining cutting-edge technology with a steadfast commitment to sustainability.
At Firmus, we are unique in our approach. We design, build, and operate a new class of digital infrastructure – the AI Factory. Through our model-to-grid technology approach, we have pushed the boundaries of multi-generational liquid cooling systems, energy management, AI software orchestration, and construction. For our customers, this approach allows us to make every watt count and deliver low-cost AI tokens globally.
Firmus AI Cloud
Our large-scale GPU cloud platform, Firmus AI Cloud, is purpose-built to deliver energy-efficient AI compute at scale to customers.
It empowers developers, enterprises, educational institutions, and government users to train and deploy AI models with unmatched efficiency and cost savings. With an ever-growing suite of services and applications, we are committed to delivering a cloud experience that is market-leading, proprietary, and built to scale.
ROLE SUMMARY
Firmus Technologies is seeking a Principal Engineer, AI Cloud Software to own the Firmus AI Cloud API and the control plane behind it. Customers, partner platforms, and AI agents use our API to provision and run GPU compute, and our console, SDKs, CLI, and Terraform provider are all built on it. You will set the technical direction for all of it. Everything you design must work across sites and scale to hundreds of thousands of GPUs in gigawatt-scale AI factories.
KEY RESPONSIBILITIES
Architecture and decisions
- Own the overall design of the public API: the resource model, contracts, versioning, and the compatibility promise customers and partner platforms build against. Set one contract style, one identity model, and one versioning policy that every domain team follows.
- Own the software architecture behind it: service boundaries and the control-plane model. Set a target architecture two to three years ahead and plan the steps from today's system to it.
- Design the control plane to keep each region serving when another fails, protect it under load with per-tenant rate limits and fair queuing, and remove the limits that would block growth in tenants, resource types, sites, and request volume.
- Lead design reviews for significant changes across the teams, sign off architecture decision records, and settle technical disputes escalated by engineers. Decide at pace with incomplete information, write down the assumptions, and revisit a decision when the evidence changes.
Build and evolve
- Write production code for the hardest parts of the API and the control plane.
- Expose the full lifecycle of compute, virtual networking, storage, and Kubernetes resources through the API, on top of the infrastructure, network and storage controllers' other teams own. Give every resource a stable identifier, track multi-step requests to completion, and offer break-fix actions such as reboot, reset, and maintenance notices.
- Build identity and tenancy into the API: organisations and projects, federated sign-on with MFA, automated user provisioning, group-based roles, and service accounts.
- Give customers, partner platforms, Firmus operations teams, and AI agents acting for a customer safe access through the same API, with scoped credentials, approval for destructive actions, and a full audit trail.
- Keep the console, SDKs, CLI, and Terraform provider consistent with the API, built from a single contract. Ship each capability as a supported feature, not a one-off for a single customer.
Reliability, security, and cost
- Set SLOs and error budgets for the API and control plane and use the budget to decide when reliability work comes first.
- Lead the technical response to major incidents in the API and control plane and make sure every post-mortem ends with a fix that stops it happening again.
- Design usage records accurate enough to bill from and for customers to verify, with quotas and audit logs that customers can query through the API and export.
- Design every service to survive rolling node replacement without customer impact, and keep dependencies and images patched within policy.
- Keep Firmus' SOC 2 Type 2 and ISO 27001 controls passing at every new site.
- Own cost at the design level and raise any design change that alters what customers consume.
Technical leadership
- Grow engineers through mentorship, design review, and code review. Hiring and performance review sit with the Engineering Manager.
- Set the software engineering standards other teams build to, covering API design, testing, code review, and service templates. Each team’s principal still owns the architecture of their domain.
- Set how the team uses AI to build software effectively and securely, with review gates and automated architecture checks that keep AI-generated code within the agreed design.
- Own the technical interviews and assessments for software engineers and hold every candidate to that bar.
Partnership and communication
- Shape the technical roadmap with product and deliver API and architecture work ahead of the features that depend on it.
- Agree ownership in writing at each boundary with platform, data, observability, security, compute, storage and network teams, and settle disagreements directly.
- Represent your team in customer and partner briefings on AI cloud software design, tenancy, reliability, or a contract change, and bring what you learn back to product.
SKILLS AND EXPERIENCE
- Bachelor's degree in computer science or a related technical field, or equivalent practical experience.
- 10+ years in software engineering, including at least 3 years at principal level or equivalent, setting the technical direction that multiple teams delivered against. You are still hands-on and intend to stay that way.
- Has designed and evolved a public API that external customers built against: the resource model and lifecycles, OpenAPI or gRPC contracts, versioning, and backward compatibility, with at least one SDK, CLI, or Terraform provider kept in step.
- Has run a multi-tenant control plane across more than one region or data centre, serving thousands of tenants or millions of API requests a day, with failure isolated per region.
- Has built identity and security into customer-facing software: federated sign-on (OIDC or SAML), user provisioning (such as SCIM), role-based access, tenant isolation, secrets handling, and dependency patching, under SOC 2 Type 2, ISO 27001, or an equivalent control set.
- Strong distributed-systems fundamentals: consistency, idempotency, retries and backpressure, long-running workflows, and the failure modes of a control plane under heavy load.
- Production Kubernetes experience beyond deploying to it, including control-plane services built on the Kubernetes API with custom resources and controllers.
- Strong in Go or Python, with infrastructure as code and CI/CD as part of daily work.
- Has run a reliability programme other engineers followed: SLOs and error budgets, change control, and incident reviews that changed the system.
- Clear written and verbal communication in English, able to explain problems, decisions, and trade-offs to engineers, executives, and customers.
- Willing to join the on-call rotation for the services you own and to travel overseas occasionally.
Bonus Points
- Background at an AI cloud or high-performance compute provider.
- Exposed virtual networking (such as VPCs, security groups, and private DNS) or GPU nodes lifecycle through a customer-facing API.
- Built usage metering, quotas, or billing for a multi-tenant control plane.
- Built hosted or virtual Kubernetes control planes, or tenant isolation for a managed Kubernetes product.
- Built customer-facing web applications (React or similar), with enough depth to keep the console consistent with the API.
- Built agent-facing tooling, such as agent-callable APIs, CLIs, or MCP servers, with safety and governance controls.
- Led a team's use of AI-assisted and agentic engineering, measured by delivery outcomes and token cost, and kept it secure with sandboxed agents, least-privilege access, and review gates.
- Certifications such as CKAD or CKA, or AWS Solutions Architect Professional.
LOCATION
Singapore or Australia
Find more English Speaking Jobs in United Kingdom on Arbeitnow