DevOps / AI Infrastructure Engineer

About Numezis

Numezis is the first AI-native Business Administration OS. We turn compliance-heavy back-office work into governed execution that scales. Our agents don't suggest — they execute, under rules, with full audit trail.

The platform is a modern cloud-native SaaS with an AI agent layer, hosted exclusively in Switzerland. We serve Swiss SMEs and fiduciary partners through a workspace architecture built for multi-entity scoping, permissions, and country-aware compliance.

We are a small, senior team shipping weekly to real customers. Swiss-made, audit-ready by design, Swiss GAAP compliant. The work is serious, the standards are high, the pace is weekly.

The role

You design and operate the infrastructure that keeps our agentic finance OS reliable, observable, secure, and cost-efficient at scale. You own the platform layer that turns experimental AI into dependable financial software — the backbone that auditors, CFOs, and FINMA-aware clients actually need.

You partner directly with AI leadership and engineering on platform decisions. This is a role for a pragmatic operator who sees LLM workloads as production systems with real SLAs, real costs, and real compliance obligations — not as lab experiments.

What you'll do

• Design and operate cloud infrastructure for AI-powered services — containerized, infrastructure-as-code, GCP Zurich multi-zones.
• Build CI/CD pipelines tailored for AI systems, including prompt, model, and agent releases, with eval regression gates.
• Implement observability for LLM and agent workloads — tracing, evaluation, drift detection, cost attribution per customer.
• Own reliability, scalability, and incident response for the agentic workflows that customers depend on daily.
• Build and maintain evaluation and regression pipelines that run on every commit — no release ships without green evals.
• Optimise inference cost, latency, and throughput across providers and selectively self-hosted models.
• Harden the stack around security, secrets management, data residency, and auditability — our clients rely on Swiss data residency as a contract term.
• Shape the developer experience so AI engineers can ship safely and fast, without operational drag.

What you bring

• Strong DevOps / SRE background — cloud (GCP preferred, AWS/Azure acceptable), Kubernetes or equivalent container orchestration, Terraform, CI/CD.
• Experience running production systems with real reliability, observability, and incident-response requirements.
• Hands-on exposure to LLMs, RAG pipelines, vector databases, or agentic systems — you understand what makes LLM workloads different from standard web workloads.
• Solid grasp of security, networking, and data-protection fundamentals.
• Engineering mindset rooted in correctness, automation, reproducibility.
• Comfort bridging infra, AI, and product concerns — you are not in a silo.
• Written communication that makes runbooks, postmortems, and eval dashboards legible.

Bonus points

• Experience with LLMOps tooling — Langfuse, Braintrust, LangSmith, Arize, Weights & Biases, OpenLLMetry.
• Experience running RAG or agentic systems in production with measured SLAs.
• Familiarity with multi-agent orchestration or agent interoperability frameworks.
• Background in fintech, ERP, or other regulated industries.
• Knowledge of Swiss and EU data residency and compliance constraints — FINMA-aware architectures, nLPD, GDPR.
• Experience operating on GCP specifically (Cloud Run, Pub/Sub, DocumentAI, Cloud SQL, Workflows).

What makes this role unique

• Build the infrastructure behind cutting-edge agentic AI applied to real financial operations.
• Define the reliability and observability standards for production AI agents in a compliance-heavy domain.
• Shape security and compliance for a Swiss enterprise SaaS product, in a market that takes data residency as a first-class requirement.
• Influence architecture, not just operate it — direct collaboration with AI and engineering leadership on every major decision.
• Join at a meaningful moment in the product's lifecycle, before scale-up bureaucracy.

Benefits & perks

• Remote-friendly setup — Geneva, Zurich, or EU timezone
• Meaningful equity for early hires, 4-year vesting with 1-year cliff
• Swiss-market-aligned base salary, transparent range at offer stage
• Generous vacation policy and deep-work culture
• Top-tier equipment of your choice
• Quarterly team offsites — Swiss Alps cadence
• Fast feedback loops: we move with intent, not bureaucracy

Our hiring process

Intentionally lightweight and respectful of your time. We aim to respond within 24 hours at every stage.

• Initial conversation — 30-minute call with leadership to align on values and role scope.
• Technical / craft assessment — a paid, real-world challenge that mirrors what you would ship in week one.
• Team interview — panel with leadership and one advisor, focused on culture and collaboration.
• Reference checks and offer — transparent compensation, equity, and onboarding plan.

Ready to apply?

We'd love to hear from you.

Contact — hello@numezis.com

Subject line — Application — DevOps / AI Infrastructure Engineer — Agentic Systems at Scale