Inspiration ...We wanted to streamline the process for students to calculate current and projected grades. What it does After taking in syllabus and schedule input from a student, the program generates an excel file that allows a student to input their scores and projected scores for each assignment throughout the semester to calculate a projected grade How we built it We decided to use React.js for the front end and Django's rest-framework for the back end. We used a variety of document processing libraries to allow us to parse through different file types in order to detect information about assignment weights and deadlines. Challenges we ran into Part of our parsing process was to churn down the document using an LLM, the output wasn't super consistent, and a large part of the issues we ran into creating the backend stemmed from this inconsistency. Accomplishments that we're proud of We're proud of building unfamiliar functions, such as transforming plaintext into an xlsx file, and following a more structured approach for development, such as beginning frontend development with a wireframe. What we learned
We learned how to better allocate time to each stage of development and how to utilize hooks in React for file transfers via API.
What's next for CourseCompiler
Next, we change from writing to plaintext .xlsx to using Google Apps Script to write directly onto a Google Sheets file that the user can directly make a copy of, which would increase efficiently as it cuts out the extra download-upload for the user. Furthermore, we would refine how Gemini detects certain information from the parsed text via restructuring the data, refining the prompt, or providing context if needed.
Log in or sign up for Devpost to join the conversation.