Consortix aims to expand its current AML solutions through machine learning methods by focusing on
1. reducing the number of false alarms, while true results kept unchanged
- by eliminating initial setup errors, and
- resiliently following time changes
by continuous analysis of data and continuous self-learning optimization of existing rules,
2. revealing new relations and developing new rules by detecting new parameters (deep learning) in
- the transaction chain and
- client profile
by linking parameter fields (using matrix factorization model).
The two years-long project is expected to deliver its first viable results by the end of 2020.
More info on the microsite connected to the project with our AI video: https://nextgenaml.co.uk/
