Syllably

• 

  Home page

• 

  Asking about the class location

• 

  Asking about the grading component

Inspiration

Ever found yourself drowning in syllabi during your course selection in your college years? Just a couple of weeks ago, I went through just that: the same old and tiring process of sweeping through every lengthy syllabus out there to find THE course that you deem good. Thinking the next semester depends on the that very research, students go through that cumbersome process semester after semester. If only there was a simple and free tool ...

What it does

Accounting for every penny, we students worship useful and free tools to help us study more efficiently. The quest of finding a study tool that won't break the bank and is easy to use is a hassle. ENTER Syllably. This simple tool that I have built will enable fellow students to easily extract the schedule written in the syllabus as a calendar file and help them chat with a generative LLM about the syllabus contents. Utilizing Gemini-Pro model and Streamlit's minimal style frontend interface, I am sure Syllably will offer an efficient solution to the common problem.

Challenges

Helping me to gain more insight and research, the gemini pro model lacked support for document processing. Thus began my quest for a PDF reading tool. Initially, I pinned my hopes on the PyPDF2 library (and I thought it was THE thing). Combining the pdf parser and regular expressions, I believed I have solved my issue. However, turns out (and obviously) that most syllabi differed greatly in formats. Plus, the library was adding unnecessary whitespaces to the extracted text, making it harder for the model to process. To add, the model was generating inconsistent contents, so I had to prompt-engineer.

What I learned

Syllably truly helped me to integrate my fragments of knowledge into a whole and practical tool. I am proud that its initial implementation has already proven its value, especially to me and to my circle of community. I'm excited about its potential to benefit students far and wide.

What's next for Syllably

I will aim to enhance Syllably by fine-tuning an appropriate model to specialize more in syllabus parsing and explanation. An immediate feature that I want to add is the capability of the chat bot to suggest personalized study schedules and relevant practice resources based on syllabus content. Furthermore, I plan to integrate assignment to-do lists, tracking features, and integration with other study tools to enhance Syllably's functionality.

Built With

• gemini-pro
• python
• streamlit