Senior DevOps / Platform Engineer

About Shelf The enterprise is going agentic and we are creating a unique operating system for an automated future. Most AI agents break the moment they hit real business complexity. That’s why we built a platform that models a company's policies, workflows, and operational logic into an AI Data Model.

Shelf enables AI Agents to reliably reason and deliver precise and reliable outcomes at scale. We already have what most AI startups are still trying to earn: Enterprise customers such as Glovo, Nespresso, and HelloFresh, real production data, and tens of thousands of users. Customers trust us with operational knowledge: Our platform is highly reliable and secure with SOC2, HIPAA-ready security practices, and mature integrations in place.

Location – Warsaw, work from the office.

About The Role

This role is for a senior DevOps / platform engineer who can own the AI platform infrastructure that lets Shelf teams ship agentic AI capabilities quickly, safely, and reliably. We want someone who is strong technically, operationally mature, communicates clearly, and has the judgment to make infrastructure simpler, safer, and easier to evolve over time. Expect real ownership from platform design through production operations.

You will own the platform layer behind model access, model deployment, observability, reliability, security, and developer experience. That includes our internal AI gateway, cloud-hosted models, self-hosted models, VLMs, audio models, and the runtime paths product teams need to build production AI systems. The team is small enough for platform engineers to make real decisions.

Career growth comes from ownership. Titles matter less than the scope you can carry. The people who grow fastest take responsibility for larger systems, unblock product teams, drive reliability, and develop better engineering practices.

You solve ambiguous platform problems by shaping the right operating model, then ship, automate, instrument, and keep it healthy in production. We build AI-native. Engineers use Codex, Claude Code, and agents they build themselves.

We invest in harness engineering through skills, CLIs, logs, traces, and orchestration. In this role, you use that same mindset to give product teams better paths for deploying, scaling, observing, and securing AI workloads. This is a role for someone in a high-growth chapter of their career, who wants to do the best work of their life, learn fast, and win as part of a team going all-in on a hard mission.

What You Will Own Own Shelf’s AI platform infrastructure end to end: model access, model deployment, runtime paths, observability, reliability, security, and developer experience. Build and operate the internal AI gateway and the platform services product teams use to ship AI capabilities. Enable teams to use cloud-hosted and self-hosted models safely, including LLMs, VLMs, audio models, and other AI runtimes.

Improve deployment, scaling, monitoring, and incident response for AI-heavy production services. Build infrastructure as code, CI/CD flows, environment setup, and paved paths that reduce developer friction. Strengthen logs, metrics, traces, dashboards, alerts, and service-level visibility so teams can understand their systems earlier.

Improve security and access control around models, secrets, permissions, auditability, and production operations. Reduce recurring toil through automation, internal tools, runbooks, and better defaults. What Strong Performance Looks Like Product teams ship AI capabilities faster because the platform gives them clear, reliable paths.

More models and runtimes become production-ready without each team inventing its own infrastructure. Teams have better visibility into their services and can catch reliability problems earlier. Deployment, observability, access control, and incident response become calmer and easier to operate.

The systems you touch become simpler, safer, and easier to evolve.