Irada holds a Research Framework Agreement with Deakin University, executed in 2024, under which research projects with A2I2 can be commissioned. We can collaborate with A2I2 researchers and engineers working on emerging AI capabilities in financial services. Their input leaves us better prepared to overcome the failure modes that recur in AI assessment: oversimplified frameworks, vendor-influenced recommendations, and generic opportunity lists that collapse under regulatory scrutiny.

Where an engagement involves questions at the frontier of applied AI in regulated lending — novel technologies, unusual data architectures, or unproven applications — we can commission research services from A2I2 under that agreement. Teams of AI researchers and engineers can investigate technical feasibility, develop prototypes, and evaluate whether approaches can work in your regulatory environment before commitment.

We provide independent assessment through three services addressing different decision stages: AI Strategy Assessment establishes organisational direction, AI Implementation Readiness creates execution blueprints, and AI Opportunity Validation evaluates specific use cases. All services deliver a defined report and deck. Where work cannot be scoped in advance, we provide it as consulting on agreed terms.

Our engagement model takes the analytical load off client teams. We conduct stakeholder interviews, review documentation, and perform the assessment work. Clients provide access to relevant stakeholders and materials but aren’t required to dedicate internal resources to analysis tasks. This approach recognises that technology and operations teams in mid-sized lenders are already stretched managing current priorities.

Our revenue comes entirely from assessment work. We don’t implement AI systems, sell vendor products, or receive downstream implementation revenue. This separation enables objective evaluation of build-versus-buy options, vendor alternatives, and implementation sequencing based solely on what serves your operational context and strategic objectives.

More about our assessment approach—independence positioning, research collaboration, methodology—is available on our Approach page.

We’ve built production-ready IFRS 9 expected credit loss models for one of Australia’s largest non-bank lenders, covering credit cards, personal loans, and motor loans across multi-brand portfolios. Model design and documentation passed external validation with no issues identified. We assess regulatory feasibility against standards we have met in production.

Through feasibility investigation conducted with A2I2, we examined AI-assisted income verification, establishing where models deliver operational value and where they create compliance obligations. Navigating regulatory interpretation, vendor constraints, and data quality limitations in production informs every feasibility assessment we provide.