Traditionally, payment cards issuers offer standardized products targeting macro-clusters of customers. Therefore, cardholders can only choose between a basic product and a premium one that comes with a prepackaged bundle of additional services. As in other industries, this one size fits all strategy is being replaced by offering a wide range of additional packages which help customers customize products. However, companies are not making the most out of big bulks of customer data available, and struggle to offer value added services that match customer’s needs and habits. Eventually, companies’ contact channels are either standard, with a massive target and low cost per contact (e.g. DEM), or supported by human interaction, which is highly valuable but too expensive for mass usage (e.g. contact center). How can companies engage and recommend their customers on a one-to-one cost-effective way leveraging on available data?
Selly chatbot represents a key contact point for engaging cardholders. It guides them towards the next-best-product/service, matching real needs through custom interaction. · Selly recommends customer which products/services to subscribe for (e.g. travel insurance policy, loyalty redemption), tweaking suggestions according historical data or real-time inputs · Selly answers customers’ requests (e.g. credit card balance) and reacts in-real-time with proper actions (e.g. spending control, card suspension) · Specific events (e.g. registration completion, over-limit) could trigger Selly proactive intervention and enrich customer journey · Interaction changes according users profile (e.g. age, gender)
Selly operates combining a number of key components: · Big data: Selly takes as one of the main inputs the data already available to the company or real-time information from users · Machine learning: technically, Selly processes the data using advanced machine learning tools · Internal services: Selly can talk with company internal services to retrieve specific information or trigger core operations
Jacopo Agnelli/Erich Fortuni (e*finance consulting) Marios Alitska (Sytel) Marco Argenti (IrisCube) Raffaele Rotella (Open) Giorgio De Donno/Giulia Grosso (Triplesense)
to start the project:
-clone the repository;
-import the project into your workspace
-download in csv format che sheet Transaction Hystory and Users from the SELLY CONCEPT Excell file
-put them within the Selly\src\main\resources folder
-Modify the ClientRequestManagerService.java class, PARAMETER --> private static final String authorization = "Bearer ##AUTH-TOKEN##"
inserting the Selly project (from Api.ai) authorization token.