Case Study: Tide achieves 97.6% invoice-to-transaction matching accuracy and saves 1,000 hours of manual labeling with Snorkel AI

Tide - Customer Case Study

Tide, a UK-based fintech, needed to match receivable invoices from its mobile app with incoming bank transactions but was relying on costly subject-matter experts to hand-label historical data. To scale and speed this matching, Tide engaged Snorkel AI to automate the labeling and data preparation process.

Snorkel AI implemented programmatic labeling using labeling functions to extract information and encode business rules, creating training data without manual efforts. The solution delivered 97.6% accuracy in detecting the transaction corresponding to a given invoice, replaced roughly 1,000 hours of hand labeling, was built and deployed in 15 days, and has been applied to about 5 million invoices.

Tide