Productive Training Using E-mail Interface

An application developer who builds user applications on an operating system like Linux spends most of his time compiling the code for traditional applications and training the model for Machine learning (ML) applications. For example, the training image classification ML algorithm takes several hours. So, any notification to the developer which provides the details of the task once it is completed and given them the facility to execute operating system commands remotely, will immensely help the developer and boost his/her productivity on a large scale idea can develop companies’ productivity. Therefore, proposed is a method, an email operating system interface in which application developer can use to get a notification when his task is completed and also use the same interface to execute any operating system commands remotely using any mobile device. The proposed solution follows the multithreaded architecture, which helps to cater to the request from multiple users at the same time. The current solution uses a python daemon process to receive the operating system commands from a simple mail transfer protocol email interface and execute the same in a separate thread and share the output back to the user. Alternatively, users can employ this tool to get mail notification once the task like training algorithm is completed and based on the output user can modify the hyper-parameters and re-issue the model training command. Thus, with this approach productivity will help software developers and especially machine learning developers to speed up their software development efforts.