Help understanding, building and deploying a neural network for activity recognition

I am trying to build a deep neural network based on this paper: [login to view URL] 1) I need help understanding the data pre-processing required based on data collection output from the sensors I have and the format of the data needed. 2) I need help understanding how to choose the necessary parameters for the specific activity recognition I am doing (window size and everything else), and answer some questions on doubts and approaches that I have brainstormed. 3) I need help understanding how to deploy this neural network in the most efficient way that will allow others to to simply download an app and upload their own data from the same sensor to continue to train the algorithm without any work needed other than uploading their data I want to schedule 2 30 minute consultations where the paper has been read beforehand, the nature of what I am trying to do has been communicated thoroughly and understood ahead of the consultation so that all the questions I have can be answered from a well-informed reference frame. I am looking for someone who is an expert in this field, who has experience writing DNNs very similar to what is asked for above, and experience with deploying it onto apps. I need a mentor/tutor, and if satisfied willing to extend the relationship long term.