AI Devops Engineer

AI DevOps Engineer

Triskele Labs is a large-scale sovereign Australian cybersecurity provider. We deliver Managed Detection and Response (MDR), Digital Forensics and Incident Response (DFIR), offensive security, and Governance, Risk and Compliance (GRC) services to some of the most regulated and high-value organisations in Australia, across financial services, government, education, healthcare, and critical infrastructure.

We operate a sovereign 24x7 SOC, a national DFIR practice, and a growing engineering function that builds the tooling, automation, and AI capability our delivery teams rely on. Engineering excellence and operational discipline are core to how we win and keep clients.

The AI DevOps Engineer is a key role at Triskele Labs. You will sit inside the DevOps team and report to the DevOps Architect, who owns the broader cloud, platform, and DevOps stack. Your mandate is to own the design, deployment, and ongoing operation of our AI and AI enablement capability across the business.

This is an AI-first role. The majority of your time will be spent building and operating the platforms that host our internal and client-facing AI capability. Primarily this means AWS Bedrock for sovereign access to frontier and open-weight models, with selective self-hosting on AWS where workload economics or specialisation justify it, alongside workflow orchestration and agentic automation, and enterprise search and retrieval.

You will also contribute to cloud and broader platform work alongside the rest of the DevOps team.

You will be hands-on, opinionated, and accountable for outcomes. Triskele Labs is a fast-moving environment where engineers are trusted to make calls, ship, and own what they build.

AI Platform and Enablement (primary focus)

• Design, deploy, and operate AWS Bedrock as the primary inference platform for sovereign internal and client-facing AI workloads, including model access, integration patterns, throughput management, and quota planning across frontier and open-weight model families.
• Design, deploy, and operate open-weight LLM hosting on AWS for sovereign and client-facing use cases, including model selection, fine-tuning workflows, and inference optimisation.
• Own the workflow orchestration and agent automation platform, including hosting, scaling, security, integrations, and enablement of business users building workflows.
• Own model selection and routing across the AI estate so every workload uses the most cost-effective model that meets the quality bar, including modelling unit economics across managed and self-hosted inference options.
• Stand up and operate enterprise search, retrieval, and embedding infrastructure, including vector search, index lifecycle, ingestion pipelines, and query performance.
• Define and enforce patterns for prompt management, evaluation, observability, guardrails, cost control, and model lifecycle across the AI estate.
• Partner with engineering, product, and delivery leaders to identify, scope, and deliver AI use cases that drive measurable business outcomes.

Engineering, Scripting, and Automation

• Write high-quality, production-grade AI-assisted code across Python, TypeScript/Node, and Bash to automate platform operations, integrate systems, and build internal tooling.
• Design and build APIs and integrations between AI services, internal business systems, and the security tooling stack.
• Apply infrastructure-as-code (Terraform preferred) and GitOps patterns across the AI and DevOps estate.
• Embed security, observability, and cost discipline into every platform and pipeline you build.

Cross-functional Contribution

• Contribute to the wider DevOps roadmap and on-call rotation as a senior member of the team.
• Document architecture, runbooks, and standards so the rest of the business can adopt and extend the platforms you build.
• Coach engineers and non-engineers on safe, effective use of AI tooling and automation.
• Stay current with the open-weight model ecosystem, hyperscaler AI service roadmaps, and the broader AI platform engineering discipline.