Senior DevSecOps Specialist, Agentic Infrastructure

oHallo is hiring a senior DevSecOps specialist to be responsible for the security and operation of the infrastructure behind our AI agents. oHallo’s agents answer customers across every channel and complete operational tasks inside each business’s own systems. Every one of those tasks is carried by a workflow that calls language models, acts in business systems and records a complete audit trail. You will keep that environment secure, reliable and observable in staging and in production.

Agentic workloads behave differently from conventional web traffic. A single customer request can unfold over several days, involve human decisions and reach into systems owned by the customer. Agents operate inside systems that hold customers’ most sensitive data, and every action they take is governed by policy. Voice conversations demand real-time responses. This role exists to secure and operate that environment with the discipline of a senior engineer and to give the team clear oversight of how every agent behaves.

You will join the engineering team and work in a cloud-native production environment. It runs on Kubernetes with GitOps deployment through tools such as ArgoCD, together with durable workflow orchestration, event streaming, relational databases and centralised monitoring and logging. The entire infrastructure is hosted in the European Union. Our customers adopt oHallo through procurement and security review, and security is part of every engineering decision we make.

Responsibilities

• You are responsible for the security and operation of the production and staging Kubernetes environments, including hardening, availability, capacity, upgrades and cost.
• You are responsible for the security of the software delivery pipeline and the integrity of everything it releases to production.
• You safeguard the infrastructure through secrets management, least-privilege access, network segmentation and control of outbound traffic.
• You lead threat modelling for AI agents across prompt injection, tool permissions, isolation between customer accounts and access to customer systems, and you review new designs against it.
• You operate the workflow orchestration and event streaming that carry every agent task, and you ensure that each deployment preserves the workflows already in progress.
• You oversee the monitoring of agent behaviour in production, including model latency, tool-call outcomes, workflow health, token consumption and anomalous access.
• You maintain service-level objectives and alerting, and you lead the response to operational and security incidents.
• You safeguard the availability of every service that agents depend on, including language model providers.
• You oversee the release process from continuous integration through staging to production, including promotion, rollback and the safe delivery of database migrations.
• You are responsible for backups, restore testing and disaster recovery across every stateful system.
• You supply the technical evidence behind customer security reviews, coordinate external penetration tests and implement the technical controls that GDPR and the EU AI Act call for.
• You participate in architectural decisions so that every new service is secure and ready for production from its first deployment.

What we are looking for

• At least eight years of experience in DevOps, security engineering, site reliability or platform engineering, with a substantial part of that time spent securing and operating production systems that customers depend on.
• Deep, hands-on experience with Kubernetes in production, including hardening, upgrades, networking, storage and recovery from failure.
• Practical experience securing the software supply chain, including container image scanning, dependency management, signed build artefacts and pipeline security.
• Strong knowledge of cloud and network security, including identity and access management, secrets management, encryption and segmentation.
• Experience with GitOps and infrastructure as code, preferably with ArgoCD.
• Proven experience operating stateful distributed systems such as relational databases, event streaming platforms and caches under real load.
• Hands-on experience operating production systems built on large language models, with an understanding of their failure modes and security risks, including provider outages, rate limits, non-deterministic output and prompt injection.
• Strong command of observability, including metrics, logs, tracing and alert design.
• The ability to read and change application code in languages such as TypeScript and Python, so that you can diagnose problems and review security-relevant changes across the full stack.
• Experience leading the response to operational and security incidents and writing clear, blameless post-incident reviews.
• The judgement to make sound decisions independently and the ability to explain those decisions clearly to engineers and to leadership.
• Professional written and spoken English, which is our working language.

Experience with durable workflow engines is a strong advantage. A certification such as the Certified Kubernetes Security Specialist (CKS) is valuable, as is working knowledge of the OWASP Top 10 for LLM Applications. Familiarity with real-time voice systems and with the major cloud providers is also useful. Knowledge of the security expectations of European customers, including GDPR, NIS2, the EU AI Act and ISO 27001, will help you from the first day.

Location

The role is based in Copenhagen with a hybrid working arrangement. Our working language is English.

How to apply

In your motivation, include a short description of the production systems you have secured and operated, and of why this role interests you.