Principal DevOps Engineer

Remote from: Europe
Annual salary: Undisclosed
Salary information is not provided for this position. Check our Salary Directory to estimate the average compensation for similar roles.
Department: DevOps & Infrastructure
Employment type: Full Time,
Job posted: 16 Jun 2026
Apply before: 16 Jul 2026
Experience level: Senior
Views / Applies: 27 / 4

About company

About Zartis

Your Compass for AI Transformation.

Computer Software
2009

Actively Hiring

AI Summary

This is a Principal DevOps Engineer role at Zartis, a global AI transformation consulting partner, working on a founding-stage initiative to build an internal agentic platform for a large digital marketplace group. The role involves designing secure agent runtimes, building evaluation pipelines, implementing observability and cost control, and creating reusable scaffolding for multi-cloud, multi-stack environments. The ideal candidate has deep experience in developer platforms, secure execution environments, and MCP-style integrations. This is a high-impact, high-autonomy position requiring strong SRE and instrumentation skills, with a focus on enabling AI-driven workflows across multiple marketplace products.

Role DNA

Job Complexity

Easy Hard

Pace & Pressure

Relaxed Fast-paced

Autonomy Level

Guided Full Ownership

Communication Load

Independent Highly Collaborative

AI Insight The role requires building a platform from scratch, deep expertise in multiple complex domains (security, observability, agentic systems), and operating at a founding stage with high ambiguity. This demands rare technical depth and problem-solving skills, justifying the highest difficulty rating.

Salary Analysis

AI Insight No salary was specified in the job listing. For a Principal DevOps Engineer in the US market, typical compensation ranges from $150,000 to $250,000 annually, with a median around $190,000. The actual offer may vary based on location, experience, and equity components, but this is a senior role commanding top-tier pay.

Key Skills

DevOps Kubernetes Docker CI/CD Infrastructure as Code Observability Security Python Cloud Computing AI/ML Infrastructure

Cover Letter Sample

Dear Hiring Team,

I am writing to express my strong interest in the Principal DevOps Engineer role at Zartis. With over a decade of experience building and scaling internal developer platforms, I am excited by the opportunity to architect the agentic platform from its inception. My background includes designing secure execution environments with sandboxing, RBAC, and audit logging, as well as integrating MCP-style tool gateways across diverse systems.

At my previous company, I led the creation of an evaluation pipeline with golden tasks and release gates that reduced agent workflow failures by 40%. I have an instrumentation-first mindset, building observability and cost-control systems that give stakeholders real-time visibility into performance and expenses. I thrive in founding-stage environments where I can define the architecture and drive it to production.

I am particularly drawn to Zartis's focus on AI-driven platforms and the opportunity to work across a multi-cloud, multi-stack estate. I am confident that my technical expertise and collaborative approach will help accelerate the adoption of agentic engineering across the marketplace group. I look forward to discussing how I can contribute to your team.

Sincerely,
[Your Name]

Possible Interview Questions

Describe your experience building an internal developer platform from scratch. What were the key architectural decisions and how did you ensure adoption across teams?

I led the creation of an internal platform that provided self-service CI/CD pipelines, environment management, and observability. Key decisions included using Kubernetes as the substrate, Terraform for infrastructure, and a service catalog with RBAC. To drive adoption, we partnered with early adopter teams, provided extensive documentation and workshops, and iterated based on feedback. We also set up SLAs and on-call rotations to build trust.

How would you design a secure agent runtime for executing untrusted AI workflows? Walk me through the security measures you would implement.

I would use sandboxing via gVisor or Firecracker micro-VMs for isolation. Implement network policies to restrict egress, secrets management with Vault, and RBAC for fine-grained access control. Audit logging all actions. For approval flows, require human review for sensitive operations. Also, rate limiting and resource quotas to prevent abuse.

Explain your approach to building an evaluation pipeline for AI agent workflows. What metrics would you track and how would you make it cost-effective?

I would define golden tasks with expected outputs, use automated graders (e.g., LLM-as-judge) and regression tests. Metrics include success rate, latency, cost per task, and failure modes. To keep costs low, use spot instances for evaluation, cache results, and run only on code changes. Also, provide a dashboard for teams to see pass/fail rates.

How do you implement observability for a multi-cloud, polyglot environment? Give examples of tools and practices you have used.

I use OpenTelemetry for traces and metrics, Prometheus for monitoring, and Grafana for dashboards. Centralized logging with Loki. For cost attribution, I label resources and use cloud cost APIs. I implement structured logging and distributed tracing across services. For agent workflows, I track token usage and step-level latency.

Describe a situation where you had to balance speed and stability in a platform initiative. How did you handle it?

At a previous job, we needed to deliver a new self-service feature quickly. I proposed launching with a limited feature set and a manual approval gate for safety. We communicated openly with users about the trade-offs. Over time, we automated approvals and added more capabilities. This allowed us to iterate fast without compromising production stability.

The company and our mission:

Zartis is a global AI transformation and technology consulting partner where talented engineers and technologists work on cutting-edge innovation. We partner with ambitious organizations to design, build, and scale technology solutions that deliver real impact.

Our teams bring deep expertise in AI-driven platforms, secure API architectures, and cloud-native engineering. You will work on meaningful projects that accelerate the adoption of advanced technologies, from strategy and discovery through to full product delivery, helping turn complex challenges into measurable outcomes.

With engineering hubs across EMEA and LATAM, and long-term partnerships in financial services, healthcare and life sciences, and energy and climate, we offer opportunities to work on projects that truly matter. Here, you will not just build technology — you will drive business impact and grow your career alongside industry leaders.

We are looking for a Principal DevOps Engineer to work on a project in the marketplace sector.

The project:

We are looking for a talented Principal DevOps Engineer to join a founding-stage initiative at one of Europe’s largest digital-native groups, building the internal agentic platform that enables safe, observable, and reusable AI-driven workflows at marketplace scale.

You will be part of a founding platform cell at the heart of an AI House, the central engine for making agentic engineering the default across all marketplace teams. The platform does not exist yet: you help define what it is. You will work across secure agent runtimes, tool gateways, evaluation pipelines, observability systems, and reusable scaffolding, all built for a heterogeneous, multi-cloud, multi-stack estate spanning products like Leboncoin, Kleinanzeigen, Marktplaats, Subito, Mobile.de, and more.

Our teammates come from a variety of backgrounds and we are committed to building an inclusive culture based on trust and innovation.

What you will do:

Build and operate secure agent runtimes with sandboxing, runtime isolation, network policies, secrets management, RBAC, and approval flows that make agents a credible part of the infrastructure.
Design and maintain the integration surface with MCP-style adapters and gateways that let agents act on source control, CI/CD, ticketing, documentation, cloud, and observability systems across all marketplace teams.
Build and own the evaluation pipeline with golden tasks, graders, regression tests, and release gates that make agentic workflow correctness measurable and cheap enough that teams actually use it.
Implement observability and cost control including traces, telemetry, token usage, cost-per-workflow, rate limits, and failure handling so stakeholders always know what is running, what it costs, and what to do when it breaks.
Create reusable scaffolding including templates, starter kits, wrapper scripts, and workflow automation that turn proven patterns from marketplace pilots into platform components every team can adopt in a day.
Partner with the AI Architect to make the reference architecture real, in code, in production, on a schedule that does not slip.

What you will bring:

A track record of shipping internal developer platforms or developer tooling that real teams depend on, with users, SLAs, and on-call rotations, not just side projects.
Hands-on experience designing and operating secure execution environments including sandboxing, runtime isolation, RBAC, secrets, and audit logging at production scale.
A history of shipping integrations across source control, CI/CD, ticketing, documentation, cloud, and observability systems across polyglot estates.
Experience building or operating tool gateways, adapters, or MCP-style integrations for agents and an understanding of the new failure modes they introduce.
An instrumentation-first mindset where traces, telemetry, cost-per-workflow, and eval pipelines are standard practice, not afterthoughts.
An SRE mindset where reliability, rollback, rate limiting, graceful degradation, and cost control are first-class concerns from day one.
A record of shipping reusable templates and starter kits that teams actually adopt rather than internal frameworks that gather dust.
Comfort across modern infrastructure including containers, cloud (AWS or equivalent), IaC, CI/CD, APIs, and scripting languages.
Experience supporting many engineers or multiple teams as internal customers, balancing opinionated defaults with inevitable special cases.
The ability to collaborate closely with security, legal, and risk stakeholders without losing momentum.

Nice to have:

Prior experience building agentic platforms or AI infrastructure at multi-team scale.
Familiarity with MCP (Model Context Protocol) integrations or equivalent agent tool gateway patterns.
Experience with multi-cloud or multi-stack platform engineering across distributed marketplace or e-commerce environments.
Exposure to cost governance tooling, token budgeting, or LLM observability platforms.

What we offer:

100% Remote Work
WFH allowance: Monthly payment as financial support for remote working.
Career Growth: We have established a career development program accessible for all employees with a 360º feedback that will help us to guide you in your career progression.
Training: For Tech training at Zartis, you have time allocated during the week at your disposal. You can request from a variety of options, such as online courses (from Pluralsight and Educative.io, for example), English classes, books, conferences, and events.
Mentoring Program: You can become a mentor in Zartis or you can receive mentorship, or both.
Zartis Wellbeing Hub (Kara Connect): A platform that provides sessions with a range of specialists, including mental health professionals, nutritionists, physiotherapists, fitness coaches, and webinars with such professionals as well.
Multicultural working environment: We organize tech events, webinars, parties, and activities to do online team-building games and contests.