DevSecOps Engineer · Azure · Kubernetes · AI Agent Orchestration

I combine experience in infrastructure, CI/CD, and automation with AI agent engineering for cloud operations.

My approach: the model proposes, deterministic policy authorizes, and execution is verified.

• Architecture: code-defined orchestration of an LLM-assisted diagnostic agent and a deterministic remediation agent on AKS, with separate identities and permissions.
• Scenario: correct an image-reference typo after policy checks and ACR validation, verify rollout health, and escalate cases outside the permitted scope.

The PoC aims to evaluate response time from failure detection to verified recovery within this controlled workflow. Test coverage and current limitations are documented in the repository.

Python · AKS · Azure Workload Identity · ACR · Azure OpenAI / AI Foundry

• Architecture: a versioned Foundry agent connected to an Azure AI Search Knowledge Base through MCP.
• Validation: the client requires retrieval for each question and checks citations against the current results before displaying an answer.

The project includes bounded conversation history, offline tests, and local deployment commands. Source validation establishes provenance; it does not guarantee that every claim follows from the cited evidence.

Python · Microsoft Foundry · Azure AI Search · MCP

• Cloud & infrastructure: Azure, Linux, Docker, Kubernetes, Ansible.
• Software delivery: Azure DevOps, Jenkins, GitLab, Harbor, Argo CD.
• Automation: Python, Groovy, Bash, YAML pipelines.

I value engineering discipline, creativity, and responsibility: clear boundaries, least privilege, and measurable outcomes.

LinkedIn · GitLab · Microsoft Learn · Credly