AML Risk API
AML Risk API
AML Risk API for Crypto Exchanges and Payment Platforms
Understanding the provenance and destination of cryptocurrency flowing through a wallet is imperative for any exchange or payment platform sensitive to compliance. An address that accepts deposits from or makes payments to disreputable networks is potentially abetting their activity, either directly or indirectly.
The Blockseer Bitscore gives a measure on the transactions connecting any bitcoin address to potentially disreputable networks. The calculations look at transactions several generations in different directions and then apply a weighting that captures the proportion of tainted cryptocurrency flowing through the transactions.
Blockseer generates a Bitscore for an address that measures its connection to potentially disreputable parties that have directly or indirectly made payments to or received payments from that address.
Wallet Risk-Flow Graph
The Blockseer Bitscore API is the product of three processes:
Clustering is the reduction of the blockchain’s topological complexity by algorithmically identifying addresses that belong to a single entity.
Address Labeling is the process of finding who owns a particular cluster. Sources are gleaned by scouring the internet, and darknet, forums, data sharing websites, and third-party data sources.
Risk Scoring measures the propagation of “tainted” currency through the cluster graph. The score assigned to a cluster is proportional to the amount of cryptocurrency flowing through it that, through a sequence of transactions, either emanate from or are deposited to clusters associated with disreputable activity.
Q3 2018: Cryptocurrency support– Bitscore is currently supported for the Bitcoin blockchain, with support for Ethereum and ERC20 tokens to be added in Summer 2018, and Litecoin and Bitcoin Cash in Fall 2018.
Q3 2018: Geolocation API– Blockseer currently has an API (in beta) that provides source/destination country data, plus a confidence level, for Bitcoin transactions. This API will be released for production in late 2018.
– Clustering using supervised and unsupervised Machine Learning
– Transaction classification using supervised Machine Learning
– Automated algorithmic labeling from web/dark web crawling
– Risk propagation using mathematical network flow algorithms
– Graphical visualization of risk neighborhood
– Source country aided by remote node deployments.