We plan to develop a web-based application to help students plan their Computing examination revision, through training of a Large Language Model (LLM) on the Singaporean Syllabus. We hope to utilise the fastest model of the ChatGPT API possible, to ensure the student The Singaporean Syllabus can be found through the Computing textbook. The website first includes a login page where users can log in via Firebase, where their data for the application is also stored.  

After the user enters and sends the inputs, the input will be inserted into a system prompt, and then the prompt will be processed by the LLM. The system prompt will detail the LLM to output its text in a specific format readable by our application, possibly a JSON object, and will be updated in the Firebase. The output will be shown in the IndividualStudyPlanView, where users can edit specific details about one of their study plans. These fields will be easily distinguishable from one another, and have clear text on what to do, to help address design criteria number 2. When done, they can press the save button to bring the plan to the StudyPlanView, where all their study plans are shown at once. They can then navigate to the different pages through the bottom bar. Every time they change a detail there, it will be updated in the Firebase to ensure connectivity. Lastly, they can access the settings, preferences and presets pages to customise the application to their liking. They will also be able to manage their account, such as signing out or deleting it. 

Though there are concerns that utilising a website may not be as effective as an application, as it is harder to navigate, we feel that utilising a website will maximise accessibility, as it can be accessed across any platform.

Developed by: three tall ppl (Abigail, Marco and Matthias)