About the role
Site reliability engineering runs the shared platform that every venture depends on: Kubernetes clusters, the data stores behind them, the service mesh, the deployment pipeline and the observability stack. We set service level objectives with each owning team, keep error budgets honest, and make sure a failure in one venture does not take down its neighbours.
The platform carries very different workloads. Frontier AI runs large training jobs that need thousands of accelerators and careful checkpointing, while Clexerl Go serves interactive collaboration where latency is felt directly by people. Autonomy & Sensing streams high-volume sensor data from the test fleet. Capacity planning, multi-region failover and cost control are the same problem seen from three directions.
We keep a blameless culture and a small, readable toolset. Changes are made in code, reviewed like product changes, and rolled out in stages, so that a bad deploy is caught by a canary rather than by a customer.
What you’ll do
- Define service level objectives and error budgets with venture teams, and use them to decide when to ship and when to stabilise.
- Operate Kubernetes clusters, the service mesh and ingress across regions, including upgrades that cause no user-visible downtime.
- Lead the response to platform-level failures, and write the post-mortem with the changes that will stop a repeat.
- Improve observability with traces, metrics and structured logs that let an engineer find a fault in minutes, not hours.
- Plan capacity for GPU and CPU fleets, including reservations, autoscaling policies and checkpoint storage growth.
- Automate toil out of the deployment pipeline, such as progressive rollouts, automated rollbacks and migration checks.
- Harden the platform against failure with chaos experiments, graceful degradation and regularly tested backup restores.
What you’ll bring
- Experience running production systems at scale, with an on-call record and a clear account of incidents you helped resolve.
- Working knowledge of Kubernetes internals, including scheduling, networking and the control loops behind operators.
- Strong Linux and networking fundamentals, from TCP behaviour and DNS through to load balancers and TLS termination.
- Proficiency in Terraform or a similar infrastructure-as-code tool, with modules shared and maintained across teams.
- Fluency in Go, Python or another systems language, used to build tooling rather than one-off scripts.
- Experience setting alerts that page on user impact rather than on every symptom a system can show.
Nice to have
- Experience operating GPU clusters, including driver upgrades, NCCL tuning and scheduling for distributed training.
- Background in database operations, such as replication, sharding or point-in-time recovery for PostgreSQL.
- Experience with multi-region failover design and the trade-offs between consistency and availability.
- Familiarity with FinOps practice and using cost data to shape architecture decisions.
Work you might take on
- 01
Introduce progressive delivery with automated canary analysis, so a regression halts a rollout before most users see it.
- 02
Build an SLO dashboard for each venture that shows error budget burn and links directly to the owning team's runbook.
- 03
Run a chaos experiment that kills a regional database replica and documents how Clexerl Go's clients recover.
- 04
Cut GPU idle time on the training cluster by building a scheduler that reclaims reserved capacity between jobs.
Compensation
$230K – $360K
The US base salary range for this role. Every offer also includes equity in Clexerl and, when it spins out, in the venture you help build. We tell you the reasoning behind your number.
Benefits
Health, fully covered
Medical, dental and vision for you and your dependants, premiums paid in full.
Two kinds of equity
Equity in Clexerl, plus a stake in the venture you help build when it spins out.
22 weeks parental leave
Fully paid, for every parent, however your family arrives.
Rest that sticks
Flexible time off with a four-week minimum we actually check on, plus a company-wide week in December.
Learning and compute
A yearly budget for courses, conferences and books, and compute credits for side projects.
401(k) with a 4% match
Matched from your first paycheque and vested immediately.
How we hire
Apply 5 minutes
Send links to what you have built and a few lines on why this role. No cover letter needed.
Intro call 30 minutes
A conversation with the hiring manager about the work, the team and what you want next.
Work sample 2–4 hours
A paid take-home or a pairing session on a real-shaped problem. You pick the format.
Studio day One day
Meet the team, go deep on your work and ours. Remote or in person.
Offer Within 5 days
A written offer with the full range, the equity and the reasoning behind both.
Opens an email with a short template. Links beat résumés.