Case Study: a top 10 US bank boosts transaction enrichment accuracy and loyalty with Ntropy

Top 10 US Bank boosts transaction accuracy to 96.75% with Ntropy

The customer, a top 10 US bank, needed to improve its payments and loyalty experience by obtaining better coverage and merchant information from its transaction data. Its internal models were making errors in over 30% of transactions, which was costly and frustrating for consumers. The bank engaged Ntropy and benchmarked its Transaction Enrichment API to solve this challenge.

Ntropy's solution correctly recognized 96.75% of all merchants and 89% of intermediaries, outperforming both the bank's internal models and human labelers. This data quality boosted the bank's entire ML pipeline for loyalty, underwriting, and fraud. The implementation also resulted in a massive cost reduction, lowering the daily expense for enriching 60M transactions from a projected $4.5M to $9k, and slashed latency by ~260x.