Job Description
Why this role exists:
The AI Context & Knowledge Engineer owns and maintains the foundational context environment and base-level context that every autonomous AI agent applies prior to executing feature-specific work. Positioned centrally within Customer Platforms, this role solves the dual challenge of duplicated effort across delivery pods and inconsistent application of core technical, privacy, and security standards across the enterprise.
By establishing a centralised, version-controlled context layer, encompassing security policies, coding standards, architecture patterns, and platform business rules, the AI Context & Knowledge Engineer provides the highest-leverage control on output quality across all delivery pods. A single correction or enhancement made within the central base context eliminates entire classes of rework across all downstream delivery teams simultaneously.
Where will you make a difference:
In this role you will:
Context Architecture and Environment Setup:
- Establish the structure, standards, tooling, and release pathways through which all base agent context is authored, stored, and consumed.
- Design retrieval pipelines and context-injection systems used across delivery pods to ground agents in high-fidelity data.
- Build retrieval and knowledge structures that render complex domain artefacts and technical documentation usable by AI agents.
- Maintain knowledge sources, specifications, domain documentation, and code context to ensure agent reliability and eliminate hallucination risks.
Base Context Management and Governance:
- Define and maintain the core technical context layer incorporating privacy policies, security standards, coding standards, and architectural patterns.
- Act as the central authority for reviewing, approving, and releasing updates to the central base context repository.
- Ensure context changes are governed with the same discipline as production code, prohibiting unreviewed or in-place modifications.
- Retire outdated context artefacts systematically to prevent stale or conflicting instructions from reaching active agents.
Domain Knowledge Capture and Conversion:
- Partner with Principal Domain Engineers and Associate Engineering Managers to extract held domain knowledge and convert it into structured, machine-readable agent context.
- Mitigate single-point-of-knowledge risks by digitising and embedding undocumented platform patterns into shared repositories.
- Establish guidelines and boundary conditions that dictate how delivery pods contribute to and extend context locally.
- Review and integrate high-value pod-level context contributions back into the central base context layer.
Release Discipline and Continuous Quality Evaluation:
- Implement version control, branching, automated testing, and deployment discipline across all prompt and context repositories.
- Design evaluation frameworks and test scenarios to measure the direct effect of context modifications on agent output quality and rework rates.
- Track key health metrics including context coverage across domains, artefact staleness, and repeat defect rates across delivery pods.
- Analyse failure modes in agent outputs to determine whether defects originate from base context, model constraints, or feature specifications.
Who will you be working with:
- Internal Delivery Teams: Partner with Principal Domain Engineers, Associate Engineering Managers, and all pod engineers to convert domain knowledge into shared context and govern local context extensions.
- Quality Assurance: Collaborate with Associate AI Agent Evaluation Engineers to diagnose root causes of output failures and refine contextual boundaries.
- Architecture Teams: Work alongside Solution Architects to feed recurring architectural decisions directly into the base context, reducing the need for human intervention.
- Governance and Enablement: Consult with Security, Privacy, Risk, Compliance, and the AI Enablement Manager to translate policy obligations into foundational context and drive adoption across delivery pods.
What Success Looks like:
- Output Quality: Measurable reduction in agent output defects, rework rates, and repeat corrections across all delivery pods.
- Context Currency: Central context repositories maintain high coverage and low staleness, keeping pace with evolving platform and security standards.
- Adoption Rate: High adoption of the central base context layer across pods, eliminating redundant, pod-specific context authoring.
- Compliance Consistency: Uniform, default application of privacy, security, and governance rules across every agent execution without manual oversight.
- Demonstrates behaviours aligned to Officeworks values, contributing to a respectful, inclusive and high-performing culture.
How will you lead:
Individual Contributor:
- Lives our Officeworks values and behaviours
- Proactively contributes to a safe working environment, escalates appropriately if there are unsafe conditions or inappropriate behaviour
- Operates in line with applicable Officeworks company policies and Code of Conduct
- Demonstrates a strong sense of personal accountability and curiosity to learn and develop
- Contributes to an inclusive and respectful environment where diverse perspectives are valued and everyone feels safe to speak up and belong
Our values shape the way we work, our behaviours and how we interact with each other, our customers, suppliers and communities.
Our Values:
We begin with customer:
- We start with our customer needs in every decision.
- We offer simple, helpful and affordable solutions.
- We listen, learn and continuously improve.
We do the right thing:
- We lead with integrity and care for our team communities and planet.
- We work together safely and treat everyone with respect, inclusion, and trust.
- We collaborate across teams and think Officeworks-wide.
We execute with accountability:
- We are accountable for our decisions, actions, results, and mistakes.
- We stay across the details and focus on delivering what matters.
- We move with pace and agility to deliver outcomes.
We lead with ambition:
- We set bold ambitions that help us to win and transform.
- We are courageous with new ideas.
We focus on transformation, not incremental steps.
Qualification and work experience:
Essential:
- Education: Tertiary degree in Computer Science, Software Engineering, or a related field.
- Experience: Minimum of 8+ years in software engineering, with demonstrated expertise in prompt/context engineering and knowledge architecture.
- Domain Mastery: Proven proficiency in retrieval-augmented generation (RAG), vector tooling, information architecture, and LLM failure modes.
- Software Engineering Discipline: Strong background in version control (Git), release management, branching strategies, and automated evaluation frameworks.
- Regulatory Translation: Working knowledge of privacy and security policies, with the ability to translate technical compliance requirements into machine-executable context.
- Stakeholder Engagement: Strong communication skills to partner with cross-functional architecture, security, and pod engineering teams across global locations.
- Quality Evaluation: Demonstrated ability to design evaluation metrics that measure the impact of context changes on output quality and defect reduction.
Preferred:
- Industry Context: Experience operating within complex retail, e-commerce, or enterprise software platform environments.
- Advanced Platforms: Hands-on experience with modern agent platforms (e.g., Claude, custom LLM orchestration frameworks) and evaluation tooling.
- Agile & Pod Delivery: Exposure to modern pod-based delivery models and scaling central platform capabilities as a service.