Site Reliability Engineer (SRE)

Mithril•Palo Alto, CA

18h•Onsite

About The Position

Mithril is an AI infrastructure platform built to make GPU compute more accessible and affordable for the world’s leading enterprises, AI startups, and the AI research community. Founded by a former Google DeepMind research scientist and Stanford CS PhD, Mithril has raised $80M across seed and Series A funding and platform revenue has grown >6x over the past year. The engineering team is lean and high-impact, and this role is a core infrastructure hire that will shape how Mithril scales its platform across a heterogeneous, multi-cloud environment. You will be a core contributor to the stability and performance of Mithril's global GPU orchestration platform, building the automation, observability, and tooling that allows Mithril to coordinate advanced compute across multiple cloud providers at scale, ensuring customers have fast, reliable access to the infrastructure they need. You will work directly with the founding team on high-impact infrastructure decisions, from SLO design to capacity orchestration, with clear visibility into the technical and commercial dynamics of the business. Unlike most SRE roles at this stage, which are primarily reactive, at Mithril, you'll also be building infrastructure that directly shapes how the marketplace operates: how supply is sourced, allocated, and monitored across providers. The systems you build will have a direct line to customer experience and company revenue, offering real ownership, a short feedback loop to leadership, and the opportunity to define how infrastructure engineering operates as the company scales.

Requirements

The profile we're hiring for combines strong systems instincts with the ability to build and own tooling end-to-end.
Candidates should be able to point to infrastructure or automation they've built that is still running in production.
3+ years of experience in SRE, Production Engineering, or Infrastructure roles at a high-growth technology company.
Hands-on Kubernetes experience: comfortable managing clusters, deployments, and troubleshooting production incidents in a multi-tenant environment.
Cloud proficiency in at least one major provider (AWS, GCP, or Azure), including practical understanding of cloud networking fundamentals (VPC, DNS, load balancing, security groups).
Coding ability: proficiency in Python or equivalent (Go, Rust, etc.) — you build tools and services, not just scripts. Willing to pick up new languages as needed.
Linux fundamentals: strong command of Linux systems, TCP/IP networking, and security best practices.
Disciplined troubleshooter: calm under pressure during production incidents, with a rigorous approach to RCA and long-term remediation.
Clear communicator: able to document processes and explain technical trade-offs to engineering and non-engineering teammates.

Nice To Haves

Experience with GPU/TPU-accelerated workloads or AI/ML infrastructure.
Exposure to multi-cloud deployments or niche/specialized cloud providers (e.g., CoreWeave, Lambda Labs, Nebius).
Familiarity with distributed systems concepts — service discovery, circuit breakers, consensus protocols.
Production experience with Prometheus, Grafana, or OpenTelemetry.
Prior experience in a high-growth startup environment where infrastructure scope expands faster than headcount.

Responsibilities

Implement and own SLIs and SLOs across Mithril's API layer and internal orchestration services to ensure customer commitments are met.
Partner with Product and Platform teams to ensure new features are designed for operability, reliability, and performance from the start.
Participate in an on-call rotation; drive root cause analysis (RCA) for production incidents and implement durable fixes to prevent recurrence.
Build and maintain dashboards, alerts, and distributed tracing within Mithril's monitoring stack to provide high-granularity visibility into our multi-cloud infrastructure.
Own the tooling that surfaces supply-side signal — GPU availability, provider health, reservation fill rates — to both engineering and operations teams.
Develop and maintain Terraform/Pulumi modules and Kubernetes configurations to manage Mithril's growing multi-cloud provider footprint.
Write clean, maintainable Python (or Go) to automate repetitive operational tasks — from provider API reconciliation to automated health checks and capacity rebalancing.
Assist in managing GPU capacity across providers, ensuring the marketplace can dynamically respond to supply fluctuations and customer demand signals.
Surface capacity constraints and failure modes early; contribute to the tooling that enables Mithril to make fast, data-driven supply decisions.