Case
Study
n.5
Type of client
Enterprise
Sector
Banking

Client requirements
System for fraud detection on stock exchange transactions
Solution
Implementation of a machine learning tool to identify fraudulent transactions in financial markets. The complexity of the functional requirement lies in the wide range of checks required and in the fact that the domain of each check is not the individual trade, but groups of trades which are not known in advance. The system raises alarm for stock/day pairs using daily stock market data and personal information as reference.
Applied technologies
Neural network
Python
Sql Server
Web Services