The idea is originated based on past experience of absence of single source of data,complexities involved in manual processes,need to automate them and use the powerful features offer by ML
What it does
The solution advises best borrowing/lending strategies , it also helps a treasurer/trader to predict unknown cash flows. This information helps a treasurer to forecast the future cash flow needs as well as a Trader to know impact on his positions so that both can plan their respective course of actions
How we built it
Using FFDC, Python(ML) and FFDC API's
Challenges we ran into
Getting data from different underlying system was biggest challenge and availability of environment to connect to underlying systems.
Accomplishments that we're proud of
Ability to predict models and testing it with real diversified data
What we learned
How Machine learning can be used for solving more complex problems, diversity of FFDC APIto do specific tasks thereby eliminating need for duplicate efforts
What's next for Centralized Cash Management
Integration with FusionCash/KTP(transaction banking) to provide as a bundled offering.Our solution can be adopted on top of their cor solutions. Increase strategy module to cover needs for trader,provide more granularity in prediction on cash flows,extension of probablistic cash flow to solve more related problems.