Unit21 Automates Risk Decisions with AI at Speed and Safely

Unit21 is redefining how companies manage risk, compliance and fraud at scale. By applying AI to core compliance workflows, the company aims to remove long-standing bottlenecks that slow growth and increase operational load.

Unit21 provides transaction monitoring, case management, regulatory filings and device intelligence. These capabilities traditionally rely on manual work from analysts and investigators. 

To tackle this, the team asked a simple question: what would happen if every job done inside the platform could be automated with AI?

Closing the gap between speed and safety

Risk teams face a familiar trade-off. As Kunal notes, onboarding customers and approving transactions at high speed comes with a degree of risk.

If organisations focus too heavily on safety, they sacrifice speed. Achieving both remains the ideal scenario.

"Ideally, you want to be in the high-speed, high safety category – the best controls in place, but at the same time allowing as many people as you can to maximise your revenue,” says Kunal.

Unit21’s approach targets the bottlenecks that slow this journey. Writing and tuning rules, identifying behavioural patterns, manually reviewing alerts and completing regulatory filings all create friction.

Automating these steps is, as Kunal describes, the equivalent of adding hundreds of people to a team.

Learning from analysts to build better rules

Human expertise sits at the heart of effective rule writing. Kunal points out that rule writers begin by speaking directly with analysts to understand where false positives originate and which rules generate unnecessary noise.

Analysts often spot patterns, such as clusters of false positives from border states or specific rules that never surface meaningful cases.

This same logic underpins Unit21’s AI-driven rule recommendation engine. The system analyses decisions made by analysts, alongside reviews completed by the AI investigation agent. 

Humans remain the final reviewers, but the agent’s conclusions feed into an iterative loop. The AI generates rules, produces alerts and continuously strengthens the organisation’s risk posture.

Building a continuous improvement cycle

Unit21’s model replicates how high-performing compliance teams refine their systems over time. 

By automating the insight gathering and rule adjustment process, the company enables risk teams to improve both speed and safety simultaneously.

Rather than slowing growth with operational friction, the platform ensures controls evolve as quickly as threats and behaviours change.

As Kunal’s examples highlight, the future of compliance will rely on frameworks that learn from human judgment and scale it through AI.

You can find out more about this topic in Kunal's latest blog post here.