AI Assistant (Leat MCP)
July 20
→
Fraud prevention & abuse controls
Leat verifies accounts, checks redemptions as they happen, and identifies patterns such as repeated claims, false referrals, and unusual points activity. Because every check uses the same customer record, activity detected through one location or channel can protect the entire program.
Bind vouchers and rewards to the customer who received them, then check every redemption across POS, e-commerce, kiosks, and other connected channels. A code used in one channel cannot be redeemed again in another, and a screenshot of another customer’s reward cannot be claimed by a different account. Single-use rules and redemption limits continue to apply even when a code has been widely shared.
Check that a referred customer is genuinely new and does not share identifiers, payment details, or a device with the referrer or an existing account. Referral rewards can remain pending until the new customer completes a qualifying purchase. Limits per customer and period prevent one person from collecting excessive referral value through repeated or coordinated accounts.
Detect patterns that may not be visible in a single interaction, including unusually fast redemptions, several accounts connected by shared identifiers, points repeatedly earned on returned purchases, or discounts concentrated around one staff account. Detection thresholds can be adjusted as normal customer behavior changes. Flagged accounts can then be reviewed, cleared, restricted, or blocked, with the reason for each decision recorded.
Contact verification at signup confirms that an email or phone number belongs to the person using it, which removes the cheapest kind of fake account before it earns anything. Flagged accounts leave the eligibility set for every incentive, so they stop receiving offers at the till, in the checkout, and from Decisioning at the same moment. Abuse rules apply on every channel because the customer record is the same one everywhere.
Flagged account reviews
Work the queue of accounts flagged for unusual patterns, and clear, restrict, or block each one with a recorded reason.
Referral verification
Hold referral rewards until the new customer is verified and has purchased, and release or refuse them on the checks.
Duplicate account detection
Find several accounts resolving to one person through shared identifiers, and merge or restrict them before welcome offers repeat.
Detection threshold adjustment
Adjust what counts as unusual redemption velocity or referral volume as the program grows, so genuine heavy users aren't caught.
Query resolution
Settle questions about incomplete referrals or rewards from a complete record.
Duplicate redemption protection
Prevent the same reward or voucher from being redeemed more than once across POS, e-commerce, kiosks, and other connected channels.
