A horse racing regulator sought to create a faster, more effective workflow that its small team of investigators could use to tackle fraud, expose criminal networks, and strengthen data integrity. The most pressing focus was to examine instances of potential collusion, while maintaining high confidence in the accuracy of their findings.

While collusion in the thoroughbred horse-racing industry is rare, ongoing vigilance against it serves as an assurance of fairness to fans and the betting public, as well as a deterrent to those who would undermine it.

A key to identifying potential collusion is understanding the complex relationships that exist between owners, trainers, jockeys and stable employees, which are often fluid and rapidly evolving. This posed a steep barrier to timely analysis and investigation. Analysts must piece together disparate and diverse information to infer who is connected to whom and in what way. Furthermore, performing such work with large-scale data creates a high technical bar, forcing subject matter experts to rely on data scientists and engineers.

The investigative team needed a tool that could:

• Enable quick, managed access to entities of interest

• Easily identify patterns such as individuals connected across cases

• Reveal multi-hop and multi-type relationships such as employer-employee, beneficial ownership, and physical access