Platform Engineer, Core Systems

Role summary

Temporal is seeking a Platform Engineer to own the core infrastructure for its low-latency systems, including Solana validator infrastructure and trading systems. This role involves developing internal Rust libraries for logging, metrics, secrets, and configuration, as well as managing a bare-metal fleet end-to-end. Responsibilities include provisioning, configuration, security, observability, and operating Solana validators. The ideal candidate has strong Rust proficiency, deep Linux and networking fundamentals, and experience with production bare-metal operations and secrets management.

Familiarity with Solana is a plus. This is a high-autonomy, high-accountability role with direct production access.

Temporal
· New York City (preferred) or US Remote
· Full-time

Temporal builds low-latency infrastructure at the core of Solana. We operate across transaction landing, validator infrastructure, and proprietary trading systems. Our flagship product is Nozomi, for transaction landing. All of it runs on bare metal, in many regions, placed near the validators rather than near a cloud region. This role owns that layer, and the code every other engineer imports to use it.

We are hiring a platform engineer to own the infrastructure everything else runs on.

Trading systems, transaction landing, validators, and the data pipelines underneath them all sit on machines that someone has to provision, configure, secure, deploy to, and watch. Every one of those services also logs, emits metrics, reads secrets, and loads configuration, and today each of them does it its own way. You would own both halves: the fleet, and the standard libraries that make every service on it behave consistently.

This is not a Kubernetes-and-tickets role. There is very little Kubernetes here. The hot path runs on a microsecond budget, so the standard platform answers are often too expensive, and the machine you are about to reboot may be about to produce a block.

• Own the internal Rust libraries every service imports: logging, metrics, tracing, secret and credential access, and configuration. Ship them with defaults that are right, ergonomics that get them adopted, and overhead the hot path can afford.

• Build policy enforcement in front of signing and key access: proxies that decide what may be signed under which conditions, fail safely, and produce an auditable record, without adding latency the trading systems will notice.

• Own the platform end to end: provisioning, configuration, secrets, delivery, observability, and on-call.

• Make host configuration reproducible across a multi-region bare-metal fleet, including the low-level tuning that latency-sensitive services depend on: CPU pinning, hugepages, IRQ affinity, kernel and NIC tuning, and clock discipline.

• Build deploy tooling that can update production without dropping customer transactions or missing slots.

• Operate Solana validators and their supporting infrastructure: snapshots, ledger and account storage, client upgrades, network configuration, and the parts of validator operation that are unforgiving about timing.

• Run and extend the metrics, logging, and alerting stack, including retention, cardinality, and cost.

• Take incidents, find the actual cause, fix it, and remove the class of problem when you can.

• Proficiency in Rust, including designing crate APIs that other people depend on. Our tooling and our services are Rust and you will be changing both.

• Experience building internal libraries or platform APIs that engineers actually adopted. You think about defaults, ergonomics, and migration paths.

• Deep Linux and networking fundamentals. You can explain what a machine is doing, from the process down to the NIC,…