In a nutshell

The task is to investigate, design, and implement a solution to enhance the Orleans cluster with capabilities for analysing customer data collected from mobile apps.

Background & motivation

Orleans is a cross-platform framework to build robust, scalable, and distributed applications.

The thesis will be carried out in the context of concrete software development in Bember A.S.. The company is involved in developing mobile apps making use of the Orleans cluster, where the concrete development is done in F#. For the company, it is important to collect and analyze data collected from the use of the apps by the customers, to help improving the apps themselves or for marketing reasons. The thesis will investigate and implement systematic ways of improving the Orleans-based solution in that respect.

Tasks

Concretely, the thesis will address the following steps.

  • A concrete specification of the data of interests, both generically as well as customer specific.
  • Exploring and discussing different techniques viable for the selected data and the needed information. To do so, two paths for analysis will be pursued, a real-time analysis, with a latency of millisecconds, and an offline analysis, with a turn-around time of hours.
    • To analyse the app data in real-time, techniques that are available in Orleans resp. F# will be transferred to the analysis of customer data. Currently, Orleans assists the programmer to analyze API data, but not customer data.
    • For the offline analysis, the techniques will include expert systems, neural networks, and other data analysis methodologies.
  • A protoypical implementation of the methods.
  • Finally, the approach will be evaluated using real or mock customer data.