Greetings!

Sprint 3 has now ended, it didn’t go as well as planed. The main reason being that all the team members had exams the second week of the sprint.

In the earlier sprints the automation team has succeeded in creating all three of the main automation modules. They do not yet fulfill all the requirements thus the goal of sprint 3 was to improve on the modules and the user experience. One improvement made to the schedule module is that is now uses Google API to create events. This limits the modules to just work on Google Calendar but it is much better then before, where it only worked on local calendar servers. We have also started on implementing the feature of deleting events that the user has added to there calendar and also on a simple GUI. Lastly we have also added some shell scripts that help with setting up the necessary environments to develop on the project.

For the next sprint the goal of the automation team is to add more features to the existing modules and to add more features that improves the users experience. It is mostly the Schedule module that needs improvements because there are a lot of fields in an event that can be filled in with information, such as location, timezone and more. Currently those fields are left blank but the goal is that as many of those fields as possible can be filled in. To improve the user experience the goal is to implement a feature that shows the user what the program has interpret the input to mean and ask for a confirmation. We also plan to have a finished version of the GUI and to be complete with the feature that deletes calendar events.

The NLP team as mostly work on refactoring the project and separating the NLP model into multiple ones that are specific for each of the automation modules. This will make it easier for us to have more specific output from the NLP models and thus the automation modules can be more complex. The new structure is that we have one model that determines which NLP model to send the input to and then the specific automation module NLP model will interpret the input. This took a lot of work so noting else has been done on the NLP side of the project.

The goal of the next sprint for the NLP team is to improve on the specific NLP models. The one that needs the most work is the NLP model for the schedule module because it has the most fields that the user should be able to fill in. This will require that both the automation team and the NLP team work together a lot but it should not be a problem, because the automation team will need to work a lot with the GUI. The NLP team will also help out the automation team a lot this sprint so that we can hopefully meet all the essential use cases we have have when were done with this next sprint.

/The RPA Tomorrow Team