Staff Engineer, Scalable AI Inference Platform

Lyceum; Zürich; Vor Ort; CHF 130.000 - 170.000; Vollzeit

Zusammenfassung

Lyceum is seeking a backend/infrastructure engineer to ensure its AI inference platform is reliable, secure and scalable as usage grows to thousands of concurrent users.

You will own scalability, reliability, and performance across the serving stack, integrate open-source inference frameworks, and evolve the infrastructure with Kubernetes, GPUs, and strong observability. Join a world-class team building sovereign AI infrastructure for Europe.

Qualifikationen

• 3+ years in backend, infra, or systems engineering.
• Proficiency in Go and Python.
• Experience operating a model serving/ML platform.
• Strong performance profiling and latency optimization.
• Familiarity with observability tooling: Prometheus, Grafana, OpenTelemetry.
• Understanding security fundamentals: network isolation, authentication, encryption, multi-tenancy.

Aufgaben

• Scalability: architect and implement systems to scale to thousands of concurrent users.
• Reliability and observability: build monitoring, alerting, and incident response tooling; design for graceful degradation and automatic recovery.
• Performance engineering: profile and optimise the full inference path for latency and throughput.
• Infrastructure evolution: evaluate/open-source inference frameworks and tooling to improve throughput, latency, or stability.

Kenntnisse

Go

Python

Model serving experience

Observability tooling

Performance optimization

Security fundamentals

Tools

Prometheus

Grafana

OpenTelemetry

Kubernetes

vLLM

Triton

TensorRT-LLM

Jobbeschreibung

Lyceum is seeking a backend/infrastructure engineer to ensure its AI inference platform is reliable, secure and scalable as usage grows to thousands of concurrent users.

You will own scalability, reliability, and performance across the serving stack, integrate open-source inference frameworks, and evolve the infrastructure with Kubernetes, GPUs, and strong observability. Join a world-class team building sovereign AI infrastructure for Europe.