Bud's models are built with a combination of natural language processing (NLP) and supervised machine learning methodologies.
Our AI models are also trained and validated using machine-assisted human labeling teams. This means Bud incurs higher material costs but allows us to deliver a peerless solution that greatly exceeds the accuracy of our nearest competitors.
We tag every transaction with three granular classification levels to help clients understand customer spending behavior at a deeper level.
We include over 200 labels within our taxonomy and achieve market-leading enrichment accuracy of >97%.
Enrich transactions with additional metadata, including clean merchant name, merchant logo, merchant payment processor, merchant website URL and locations.
This reduces the volume of disputed transactions and also enables users to visualize how much is spent on specific merchants.
We match transactions to our expansive global merchant database covering >99% of US merchants including 19.2 million US Merchants with locations.
Bud's data enrichment service detects transactions that reoccur over a daily, weekly, bi-weekly, monthly or quarterly period, including direct debits, standing orders, regular credit card subscriptions and regular transfers.
This service identifies both incoming and outgoing transactions and can identify regular transactions even when the transaction characteristics differ from period to period.
This feature reveals a customer's spending habits and helps power more advanced pattern recognition services (see Engage and Assess for further features).
