Last year, hackers, mainly from North Korea, stole a whopping $3.8 billion in cryptocurrency, making it the worst year ever for crypto theft, according to experts at Chainalysis, a company that tracks cybercrime. This is a significant jump from the $3.3 billion stolen in 2021. A secret United Nations report also revealed that North Korea stole more cryptocurrency in 2022 than in any other year. Most of the money that was lost comes from decentralized finance, but almost anyone can be a victim of crypto scams. 

Amid the COVID-19 pandemic, American investors directed millions into cryptocurrencies such as bitcoin, ether, and dogecoin, anticipating substantial profits. However, a subset of these investors experienced financial setbacks as hackers targeted their digital wallets on platforms lacking robust cybersecurity measures. 

However, Harpie, a security company supported by Coinbase Ventures and Dragonfly Capital, aims to make a difference in this situation. 

Harpie is equipped with an address scanner designed to notify users about potential phishing, cybercrime, or theft attempts, with the aim of thwarting even the most advanced cyber criminals. In a significant development last October, the company introduced a crypto wallet monitoring service that provides instant alerts for suspicious transactions, serving as a proactive measure to prevent unauthorized transfers. 

Daniel Chong, co-founder of Harpie, explains the mechanism behind their theft and scam prevention. When a user initiates a transaction from their wallet, Harpie intervenes if any suspicious activity is detected. Drawing a parallel with how banks halt transactions when detecting unusual behavior on credit cards, Harpie adopts a similar approach, meticulously scanning each transaction leaving the user's wallet. 

The assessment is swift, typically taking around 300 milliseconds. This quick response time enables us to temporarily halt transactions before they are sent to the blockchain, allowing us to conduct thorough checks and prevent any potential issues. 

Daniel further said that the platform has a sophisticated system that quickly determines if a transaction seems suspicious or not by analyzing the information attached to it. Our advanced engine allows us to specifically block only transactions that appear to be malicious. This engine has two components: One identifies addresses that are confirmed to be good. Second flags addresses that are confirmed to be bad. 

“We maintain a database of approximately one million verified good addresses, which are essentially addresses we have approved or whitelisted for secure transactions”, Daniel further added. 

Additionally, the platform possesses another engine dedicated to searching for negative indicators associated with each Ethereum address. Essentially, this functions as our blacklist. 

Further, Denial reported that in terms of databases, the primary resources include the OFAC sanction lists, which we regularly consult. Additionally, we leverage publicly available police reports to gather relevant data, specifically focusing on wallet addresses. 

Distinguishing between regular users and potential scammers is facilitated by analyzing distinct transaction histories and on-chain behaviors. The approach involves utilizing advanced big data models to make these determinations effectively.