Fraud prevention is an exercise in competing risks. A company that detects too little abuse loses money and trust. A company that blocks too aggressively creates friction for legitimate customers.
Henry LeGard's path to founding Verisoul combines analytical problem-solving, competitive sports, and startup building. Those experiences shaped a practical view of fraud: the technology matters, but the real job is making better decisions under uncertainty.
Fraud is a business problem before it is a model problem
Technical teams can measure suspicious behavior, identity signals, and anomalous activity. The company still has to decide how much evidence is enough to challenge or block a user.
That threshold depends on business context. The cost of a false positive, the value of the transaction, the customer experience, and the prevalence of abuse all matter.
A useful fraud system therefore needs a clear operating objective. It should not merely maximize the number of risks detected. It should reduce harmful activity while preserving legitimate growth.
Competitive drive helps, but discipline wins
Henry's athletic background, including competitive ultimate frisbee, developed a tolerance for pressure and an instinct to compete. Startup building rewards those traits, but only when they are paired with discipline.
Fraud changes. Attackers respond to controls, new channels create new vulnerabilities, and growth changes the economics of prevention. Founders cannot rely on one successful model or one burst of effort.
The work requires repeated learning: observe behavior, test a response, measure the consequences, and adjust without losing sight of the customer.
Trust must be designed on both sides
Fraud prevention protects a company from malicious users. It also affects the trust legitimate users place in the company.
A customer who is wrongly blocked may never see the sophisticated risk model behind the decision. They experience a company that failed them. That makes explanation, recovery paths, and human review part of the product.
Strong prevention systems make it possible to answer why a decision occurred and what evidence would change it. Black-box certainty is especially dangerous when the model can damage a real customer's relationship.
Startups operate with incomplete information
Founders rarely receive a complete dataset before they need to act. They make product, hiring, and market decisions while evidence is still developing.
Henry's analytical background helps structure those decisions, but rigor does not mean waiting for certainty. It means making assumptions explicit, choosing a test, and updating the decision when new evidence arrives.
That same operating method applies to revenue planning. A team should distinguish known facts, reasonable estimates, and unsupported preferences. The categories will not eliminate risk, but they will make it easier to learn.
Technical rigor needs human judgment
Fraud systems can surface patterns at a scale people cannot match. Humans still contribute context, especially around unusual but legitimate behavior and the tradeoffs a company is willing to accept.
The strongest design assigns each side the work it does best. Technology monitors continuously and applies consistent analysis. People handle ambiguous exceptions, policy decisions, and customer recovery.
That boundary should be deliberate rather than discovered after an escalation.
The practical lesson for operators
Teams managing risk should define:
- The harm they are trying to prevent.
- The acceptable cost of false positives.
- The evidence required for each action.
- The review path for ambiguous cases.
- The feedback used to improve the system.
The purpose of fraud prevention is not to make every decision maximally restrictive. It is to help the business grow with a level of trust and risk it can defend.
About the guest
Henry LeGard is Co-Founder and CEO of Verisoul. His background combines mathematics, competitive athletics, technical problem-solving, and startup leadership in fraud prevention.
