Fin. services
Financial services
Agent-assisted underwriting, fraud-adjacent pattern detection, regulated-data pipelines, compliance-grade retrieval.
We design and deliver custom AI systems for organisations where the cost of getting it wrong is real. We take engagements end to end: strategy, architecture, data, model, evaluation, deployment, operations.
No pyramids. No subcontractor drift. A highlly qualified and capable team of senior practitioners, plus your people, for the life of the engagement.
Agent-assisted underwriting, fraud-adjacent pattern detection, regulated-data pipelines, compliance-grade retrieval.
AI assurance and evaluation programmes, procurement support, policy-grade technical advice, sovereign deployment architectures.
Clinical decision support that is auditable, evidence-aligned, and comfortable with the reality of Australian data governance.
Forecasting, optimisation, and operator-assist tooling where downtime is expensive and explainability is non-negotiable.
Evaluation, red-teaming, and the unglamorous engineering of systems that must keep working under adversarial conditions.
We turn an ambition into a testable problem, with clear success criteria and a realistic view of data, risk, and cost.
A small joint team delivers a working system, with evaluation baked in from day one.
We pressure-test, document, and hand over. Deployment, observability, and runbooks are part of the deliverable.
Where useful, we stay on as the operating partner. Where not, we leave behind a system your team can own.
Our research programme is small, focused, and organised around the gap between what AI systems can do and what institutions can responsibly deploy.
Everything we publish is indexed on this site.
We take on a small number of co-funded research collaborations per year.
Research positions are posted on LinkedIn when we are hiring.
Tell us what you are trying to do, why it matters, and what you have already tried.