AI Engineer

Blackroc; City of Brisbane; Hybrid; AUD 120,000 - 180,000; Full time

Job summary

BLACKROC is recruiting an AI Engineer to lead the build capability in a small innovation team, delivering LLM-based agents and retrieval-grounded assistants for a Queensland Government department. You’ll drive proof-of-concept to production, focusing on Copilot Studio, Power Platform and enterprise integration.

Based in Brisbane with hybrid work (2 days WFH), the role emphasizes assurance, governance and risk controls, including FAIRA and DLP considerations, while collaborating across

Qualifications

• Proven delivery of LLM-based solutions into pilot or production in enterprise or government environments.
• Hands-on build experience with Copilot Studio, Power Platform and Microsoft 365 Copilot extensibility.
• Experience with embeddings, vector/hybrid search and chunking strategies.
• Identity and permission design in Microsoft Entra ID and DLP considerations for AI.
• AI evaluation including prompt-injection testing and release gates.

Responsibilities

• Design, build and release AI agents and retrieval-grounded assistants using Copilot Studio, SharePoint agents, Microsoft 365 Copilot extensibility and the Power Platform
• Build retrieval solutions over organisational content, including source selection, chunking, indexing, retrieval tuning and grounding controls
• Move proofs of concept to a supportable state, covering environment placement, ownership, monitoring, cost and a documented support model
• Integrate AI capability with enterprise systems through APIs, connectors and managed data flows
• Design identity, permission and data-protection controls, including least privilege, sensitivity labelling and oversharing controls
• Define and run evaluation for each solution, covering accuracy, groundedness, prompt-injection testing, regression testing and release gates
• Author the technical content of assurance artefacts, including FAIRA, security and privacy assessments and Architecture Review Board submissions
• Assess third-party AI tools and build automation to speed up assessment at volume
• Advise on self-hosted or offline model deployments where cloud isn’t appropriate
• Lift capability through reference implementations, reusable patterns, guidance and workshops

Skills

LLM-based solutions

Copilot Studio & Power Platform

Microsoft 365 Copilot extensibility

Retrieval-augmented architecture

Identity & data protection design

AI evaluation & release gates

Education

Degree in CS / Software Engineering / Data Science or related field

Tools

Copilot Studio

Power Platform

Azure OpenAI

Azure DevOps / GitHub

SharePoint connectors

Entra ID / Purview / DLP tooling

Job description

Company Overview

BLACKROC is partnering with a large Queensland Government department to recruit an AI Engineer.

Role Purpose

You’ll be the main technical build capability in a small innovation team leading AI adoption across a large, complex organisation. You’ll take AI ideas from intake, through proof of concept, to supportable production solutions, and produce the technical evidence they need to clear government assurance. The focus is on Microsoft-based agent and assistant engineering, integration and assurance rather than model development. Success is measured by solutions that go live with an owner, a support model and a defensible risk position.

Key Responsibilities

• Design, build and release AI agents and retrieval-grounded assistants using Copilot Studio, SharePoint agents, Microsoft 365 Copilot extensibility and the Power Platform
• Build retrieval solutions over organisational content, including source selection, chunking, indexing, retrieval tuning and grounding controls
• Move proofs of concept to a supportable state, covering environment placement, ownership, monitoring, cost and a documented support model
• Integrate AI capability with enterprise systems through APIs, connectors and managed data flows
• Design identity, permission and data-protection controls, including least privilege, sensitivity labelling and oversharing controls
• Define and run evaluation for each solution, covering accuracy, groundedness, prompt-injection testing, regression testing and release gates
• Author the technical content of assurance artefacts, including FAIRA, security and privacy assessments and Architecture Review Board submissions
• Assess third-party AI tools and build automation to speed up assessment at volume
• Advise on self-hosted or offline model deployments where cloud isn’t appropriate
• Lift capability through reference implementations, reusable patterns, guidance and workshops

AI Engineer

12 Month Contract

Location

Brisbane CBD

Work Arrangements

Hybrid, 2 days WFH

Required Skills & Experience

• Proven delivery of LLM-based solutions into pilot or production in an enterprise or government environment, including at least one multi-step agent or retrieval-grounded assistant used by real users
• Hands‑on build experience with Copilot Studio, the Power Platform and Microsoft 365 Copilot extensibility, including SharePoint or Graph connectors and environment and DLP configuration
• Practical retrieval architecture experience with embeddings, vector and hybrid search and chunking strategy, and the ability to diagnose and fix poor retrieval
• Identity and permission design in Microsoft Entra ID, with working knowledge of Purview sensitivity labelling and DLP as they apply to AI
• AI evaluation experience, including success criteria, evaluation sets, hallucination and prompt‑injection testing and release gates

Preferred Skills & Qualifications

• Python for integration, orchestration, automation and evaluation tooling
• Authorship of assurance evidence that has passed formal review, such as security, privacy or architecture submissions
• Deployment of open‑weight or quantised models on self‑hosted or isolated infrastructure
• Knowledge of the Queensland Government assurance context (FAIRA, IS18, QGEA AI governance policy, business impact levels)
• Experience with Azure AI Foundry, Azure OpenAI and Azure AI Search, plus Azure DevOps or GitHub
• A degree in computer science, software engineering, data science or a related field, or equivalent experience
• Microsoft certifications in AI, Power Platform or security and identity

Please note a clear Criminal History Check is required for this role.

At BLACKROC Recruitment, we are committed to fostering diversity, equity, and inclusion. We welcome applications from individuals of diverse backgrounds, including those with lived experiences. If you require any adjustments during the recruitment process, please don't hesitate to reach out to the above-mentioned BLACKROC Consultant.