The opportunity at a glance
IGO wanted to preserve organisational memory, make critical knowledge developed through years of operating assets easier to access, and support more consistent ways of working across the business. Arinco partnered with IGO to design and deliver Ask Igi, an AI-powered assistant that connects employees with approved enterprise-wide playbooks containing IGO’s policies, procedures and work instructions. Built using Azure AI technologies, Ask Igi combines governed knowledge access, user feedback and human escalation to support IGO’s broader approach to knowledge sharing and AI adoption.
About IGO
IGO is an ASX-listed Australian resources company focused on discovering, developing and delivering the critical materials needed for electrification. Its activities span critical minerals mining, exploration and development, with over 650 employees operating across multiple locations.
Business challenge
As IGO expanded its use of formal playbooks across the organisation, it recognised that the business-critical knowledge captured within them needed to be easier for employees to find and apply at scale. The playbooks provide structured guidance on standard policies, procedures and work instructions, helping teams follow consistent ways of working and improve safety and productivity.
While IGO has an established document management system, feedback from employees suggested information was difficult to find quickly, people were unsure what to look for and there was a preference to rely on subject matter experts (SMEs). While these experts played an important role in supporting teams, relying on individual knowledge meant work was not always done consistently or to the standards required.
IGO wanted a scalable way to make approved knowledge easier to access while keeping people involved when judgement or expertise was required. It also needed an approach that could support ongoing governance and improvement as its policies, processes, work instructions and broader AI strategy evolved.
Solution
IGO partnered with Arinco to design and deliver Ask Igi, an AI-powered assistant built using Azure AI Foundry and Azure AI Search, integrated with IGO’s SharePoint playbooks. Ask Igi provides employees with a single entry point for accessing approved organisational knowledge while retaining human ownership of the underlying content.
The solution was designed with governance, feedback and human oversight built into how knowledge is accessed and improved. Responses are designed to draw from and cite relevant source material in IGO’s approved playbooks, helping employees trace information back to its source.
Continuous improvement built-in
Arinco designed Ask Igi to capture user feedback and identify opportunities to improve search performance, prompts and underlying content.
Behind the scenes, a self-improvement agent analyses interactions, including the user’s question, retrieved documents, citations and the final response. Based on this analysis, it identifies areas for improvement, such as search optimisation opportunities, prompt refinements and gaps or inconsistencies in the playbooks. Recommendations are then reviewed and approved by IGO, ensuring people remain in control of all changes.
Ongoing prompt and model management
Arinco designed Ask Igi to evolve as AI technology advances, giving IGO the flexibility to assess new models, monitor performance and adapt the solution over time. Prompt and model evaluation form part of this ongoing approach, allowing the solution to evolve as requirements and technology change.
Thoughtful human escalation
When Ask Igi cannot confidently answer a question, it escalates the query to internal SMEs via Microsoft Teams. The escalation includes context and a warm introduction, so the expert understands the question and what has already been attempted. If no clear SME is identified, the request is rerouted internally to ensure follow-up.
SharePoint as the source of truth
Arinco retained SharePoint as the authoritative home of IGO’s playbooks, preserving human ownership and enabling continuous content evolution. To improve discoverability, the team incorporated semantic search and a structured glossary, enabling both Ask Igi and employees to locate relevant information without relying on internal jargon or institutional memory.
Ask Igi is accessed through a dedicated web application, and integrates with Microsoft Teams to support escalation and improvement workflows, ensuring the assistant fits naturally within IGO’s existing collaboration environment. The architecture also supports future integration into Teams and other internal portals as adoption grows.
Outcomes
Ask Igi has established a governed approach for making IGO’s enterprise and operational knowledge quicker and easier to access while keeping source content, human expertise and continuous improvement connected. As rollout progresses, early outcomes and capabilities include:
- Faster knowledge access: Employees across operations and corporate teams can quickly locate accurate, approved information.
- Early usage baseline: In July 2026, almost 80 unique users generated over 600 messages, providing IGO with an initial view of how Ask Igi is being used as rollout continues.
- Preserved organisational knowledge: Employees have been able to locate governance documents where previous owners had left the business and the document names were no longer known.
- Improved consistency: All responses are grounded in official playbooks, reinforcing standardised ways of working.
- Reduced SME dependency: Early usage patterns show reduced reliance on SMEs for routine queries, enabling them to focus on higher-value activities.
- Scalable AI capability: IGO now has a secure, Azure-based architecture with embedded agentic and feedback-driven improvement workflows.
- Increased workforce confidence: Initial feedback highlights growing confidence among employees who can access trusted answers on demand.
Internal stakeholders have described Ask Igi as “elevating the project beyond simple document retrieval” and “a standout example of AI–human collaboration”.
What's next
IGO and Arinco have established the foundations for a multi-stage AI journey. Future enhancements may include personalised responses based on role and location, broader integration with business systems and additional workflow automation. As playbook coverage expands, Ask Igi will continue evolving as an organisational capability supporting knowledge sharing, governance and operational excellence.
