WP Leader 5 – Data analytics Enablers for Financial and Insurance Services
28 November 2019, Konstantinos Perakis, Ubitech, Greece
Within the context of WP5, the consortium partners aim at providing the tools and mechanisms necessary for executing analytics over the INFINITECH data management infrastructure. More specifically, WP5 aims at kickstarting its activities with collecting and organizing the datasets needed for the training and validation of the analytics algorithms. To do so, the partners will prepare and circulate a common template to be used by all pilot partners so as to document the details of all open and proprietary datasets that will be used throughout the demonstration scenarios.
From the algorithmic and analytics perspective, the partners will be engaged with parallelizing popular incremental analytics algorithms that are frequently used in financial and insurance applications, distributing them in multiple computing nodes in order to boost their performance. The partners will also provide a framework facilitating declarative Real-Time analytics as a feature of the data analytics engine, extending SQL in order to support the declaration, execution and configuration of queries over incremental algorithms. In addition, the partners collaborating in the context of WP5 will provide a library of ML/DL algorithms that are popular and suitable for finance and insurance applications, including the algorithms that will be exploited as part of the project’s use cases, as well as simpler algorithms that FinTech/ InsuranceTech enterprises could use for testing, training and experimentation.
Last but not least, the consortium partners will specify, implement and expose Open APIs for executing the aforementioned analytics over the INFINITECH data management infrastructure, providing the means for accessing the INFINITECH datasets and algorithms in an integrated way and from a single entry point.