Plataforma para la comunicación entre clientes y asesores en una casa de bolsa utilizando algoritmos de clusterización

Effective communication between clients and financial advisors is essential when trading financial products. Due to the lack of technological tools and the time taken by traditional means, this has a negative impact on the brokerage firm’s operations. This development aims to create a web and mobile platform to establish an efficient communication channel in which the client is able to visualize transactions, operations, portfolio and profit; analyzing and authorizing the recommendations made by the financial advisor, and executing transactions automatically. Agile methodology which is based off sprints, was used for both developments, which helped us generate the necessary deliverables in each sprint guaranteeing the software quality. By applying Agile, the application requirements were met. Likewise, a three-layer architecture was chosen for the system, where React JS and React Native were the frameworks used for the communication channel development. Additionally, in order to improve the experience of trading stocks, a machine learning search engine was included. An in-depth analysis for each algorithm was performed, leading to the selection of affinity propagation as the clustering algorithm for financial symbols since the algorithm defines the number of groups. On the other hand, k-means was the chosen algorithm for grouping clients, taking in consideration the stakeholders specifications, who asked for four clusters. As a result, it is possible to improve customer service by increasing clients’ operability and reducing response time. Operation in the platforms is limited to Google Chrome version 76+, Android 4.0+ and iOS 9.0+. It is important to mention that the solution was designed for a specific user profile, adjusted to the needs of a specific brokerage firm.