Senior DevOps Engineer

About INDEV Software

INDEV Software S.A. is a technology partner to Greek systemic banks and financial institutions. We build, deploy, and operate systems that clients depend on. Our Sovereign AI platform lets banks deploy, observe, evaluate, fine-tune, and manage local and frontier LLMs inside their own infrastructure, in line with DORA and the EU AI Act. Alongside it we deliver RegTech and compliance tooling, PropTech, voice agents, payments and cards back office, and open banking APIs. INDEV holds ISO 27001, ISO 22301, and PCI DSS certifications.

The role

You will run the infrastructure behind these systems, from our own SaaS products in the cloud to on-premises deployments inside bank networks. A large and growing share of that work is the sovereign AI stack: GPU servers, LLM serving, and the services around them. You will join our DevOps team, work closely with our engineering team on a continuous delivery cycle with frequent releases, and use LLM agents as part of how you work every day. You report to the CTO.

What you'll do

Sovereign AI infrastructure

• Run GPU servers end to end: drivers and CUDA, GPU scheduling on Kubernetes, capacity and utilisation.

• Deploy and operate LLM serving (vLLM, SGLang, llama.cpp, or similar), including quantised and multi-model setups.

• Operate routing between local models and frontier APIs inside the client perimeter, with key management, quotas, and fallbacks.

• Own LLM observability (tracing, latency, token usage, and cost) and wire evaluation runs into CI.

• Deliver models and inference stacks into bank networks, including offline and air-gapped installations.

• Operate guardrails and AI security testing tooling against prompt injection and data leakage.

• Run the infrastructure for RAG workloads: vector stores and document ingestion services.

• Support fine-tuning and evaluation work with reproducible environments and model artifact management.

Platform and delivery

• Operate and improve production Kubernetes clusters, including Azure AKS.

• Manage infrastructure as code with Terraform and Ansible across cloud and on-premises environments.

• Build and maintain CI/CD pipelines in Bitbucket Pipelines for Python and .NET applications.

• Own monitoring, logging, alerting, and error tracking, and take part in incident response and post-incident reviews.

• Deploy into client bank environments under their security and change-management rules.

• Implement controls and produce evidence for ISO 27001, ISO 22301, PCI DSS, and DORA: access management, backups, disaster recovery, patching, and vulnerability management.

• Write runbooks and documentation that others on the team can follow.

• Mentor engineers and review infrastructure changes.

How you'll work with AI

• Use agentic coding tools (Claude Code, Cursor, Codex, or similar) on infrastructure code, with the same review discipline you apply to human changes.

• Write and maintain reusable agent configuration: skills, project instruction files, custom commands, and MCP servers.

• Build LLM-driven automated audits: Terraform, Helm, and Ansible review, secrets and configuration sweeps, and compliance evidence collection.

• Build agents that act on infrastructure with approval gates, least privilege, and audit trails.

Must have

• 5+ years in DevOps, SRE, or infrastructure engineering, with production systems you were responsible for.

• Strong Linux administration and networking: DNS, TLS, load balancing, firewalls, VPNs.

• Production experience with Kubernetes and containers.

• Infrastructure as code with Terraform and configuration management with Ansible.

• Experience with at least one major cloud, preferably Azure.

• CI/CD pipeline design and maintenance (Bitbucket Pipelines, GitHub Actions, GitLab CI, or similar).

• Scripting in Bash and Python.

• Monitoring and observability tooling (Prometheus, Grafana, ELK/Loki, Sentry, or similar).

• Regular, hands-on use of LLM agents in your engineering work, and at least one thing you built with them that you can walk us through: a skill, an agent, or an automated audit.

• A clear understanding of where LLM agents fail: hallucinated commands, injection through tool output, and data leaving the perimeter.

• Fluent written and verbal communication in Greek and English.

Strong plus

• Hosting or operating GPU workloads and LLM inference in production (vLLM, SGLang, or similar).

• LLM observability and evaluation tooling (Langfuse or similar).

• Experience in banking, fintech, or another regulated environment.

If you have strong DevOps fundamentals and real LLM tooling experience but have not yet run inference in production, we still want to hear from you.

Nice to have

• Supporting fine-tuning or model evaluation pipelines.

• Working knowledge of ISO 27001, PCI DSS, or DORA requirements.

• AI security tooling (Promptfoo, PyRIT, LLM Guard, or similar).

• Windows Server and IIS for .NET applications.

• Cloudflare, HashiCorp Vault, or other secrets management.

Working arrangement

Hybrid, with 8 days per month on-site at our office in Agia Paraskevi, Attica.

What we offer

• Growth opportunities

• Private health insurance

• Expense card

• Training and upskilling budget, including certifications

• Salary range: Between 45.000€ - 55.000€ (gross), depending on qualifications and experience