AI DevOps / Infrastructure / Optimisation

Infomaniak is seeking a Mid-level AI DevOps Engineer to design, implement, and optimize AI agents and LLMs on Kubernetes within its Swiss data centers.

The company is building an open-source AI platform and deploying large-scale language models to power its products. This position focuses on ensuring the quality, reliability, and user experience of these intelligent agents while managing the underlying infrastructure.

Responsibilities

• Deploy, maintain, and optimize large language models on Kubernetes, focusing on maximizing GPU and compute resource efficiency.
• Improve and industrialize GitLab CI pipelines for AI models and manage deployments using Flux CD (GitOps).
• Enhance the monitoring stack (Prometheus, Grafana, Victoria Metrics) to gain fine-grained visibility into AI service performance, GPU consumption, latency, and health.
• Drive cost and performance efficiency through autoscaling, scheduling, quota management, and image optimization.
• Ensure the robustness, security, and reproducibility of deployments in a critical, high-availability environment.

Requirements

• Strong expertise in modern LLM serving frameworks such as vLLM, TGI, or TensorRT-LLM.
• Proficiency with GitLab CI, including pipelines, runners, variables, and Kubernetes integration.
• Proven experience with Kubernetes, covering operators, Helm, CRDs, networking, and autoscaling.
• Hands-on experience with Flux CD for GitOps, HelmReleases, Kustomize, and deployment automation.
• Solid knowledge of Prometheus and Grafana for dashboards, alerting, and exporters.
• Understanding of GPU infrastructure, including NVIDIA hardware, CUDA, GPU scheduling, and monitoring.
• Ability to operate in a critical environment with high SLA and availability requirements.
• Strong collaboration skills with ML and Dev teams.

Nice to have

• Technical curiosity and a passion for innovative challenges and optimization.
• Experience with open-source contributions or side projects.
• A collaborative team player with positive communication skills.
• Humor, flexibility, and team spirit to contribute to a fun work environment.

What the company offers

• All working hours are compensated; overtime is taken back, and paternity leave lasts three weeks.
• Up to two days of remote work are available for added flexibility.
• Regular team outings to discover new activities and bond with colleagues.
• A travel bonus of up to CHF 1500 per year, depending on your commute method.
• During illness, Infomaniak provides 100% of your salary, exceeding legal requirements.
• An ethical pension fund with uncapped LPP salaries, exceeding legal retirement benefits.
• Coverage under Private Supplementary Accident Insurance (LAAC).

About the company

Infomaniak fosters a close-knit team environment where collaboration and a fun atmosphere are prioritized. The company is committed to developing an open-source AI platform hosted in its own Swiss data centers, focusing on deploying large-scale language models and intelligent agents for its products.

• Close-knit teams are considered essential to the company's success.
• The work environment is designed to be fun and engaging.