SLack Off | Devpost

Inspiration

Can't pay attention to class? Wish there was an easier way to catch up? We've all been there. As students, we understand how difficult it is to grasp content, take notes, or even stay focused in long lectures. You zone out for 2 minutes and you miss a whole unit.

What it does

"SLack Off" doesn't just summarize lectures; it revolutionizes your learning journey with a personalized touch. Unlike other AI lecture notetakers, it doesn't just regurgitate information – it provides you personalized study plans, summaries of powerpoints and lecture audio, and generates relevant practice problems based on your class material -- so you don't have to zone out or be confused in class again. Instead, you can take learning into your own hands with the guidance of a companion -- allowing you to "slack off" without sacrificing learning!

Features:

1. Efficiency Meets Effectiveness: Experience education tailored to your needs. SLack OFF doesn't simply hand you reports and answers; it breaks down your specific lecture content into an easily readable summary to make learning more digestible. Plus, you can learn at your own pace -- fast-forwarding through the comfortable stuff.

2.. True Relevance Unlike Other AI Generators: We go beyond generic summaries. Instead, our AI identifies keywords from your uploaded content to highlight important concepts and definitions. Additionally, it tailors your practice to hands-on problems or more conceptual questions, depending on whether the subject is more STEM-related or humanities-related.

3. Personalized Study Plans: We're not just another AI tool; we're your intelligent study companion, ensuring that your study sessions are not just productive but genuinely impactful. Effective learning starts with active recall, so we start with a relevant summary, followed by explanations highlighting key concepts and definitions, and finally, personalized practice to apply your knowledge. Identify your strengths, address your weaknesses, and optimize your study time efficiently.

How we built it

We used HTML, CSS, Typescript, and React for the frontend, designed/prototyped with Figma first. In the backend, we used python, Flask, and Cohere and OpenAI APIs.

Design choices

We chose a range of deep and bright blues because these shades encourage better mind flow and performance. Stronger shades of blue support the brain’s thought processes, while lighter shades help improve concentration. We hope to promote a relaxing atmosphere to study or work in -- stimulating a better state of mind, so you can keep the ideas and focus flowing.

Challenges we ran into

4 people working on the same file, using react webhooks, git configurations, and debugging all of our different code styles.
We weren't too familiar with the front end, so it was difficult to use React webhooks for the first time.
Problems with git
(Also, Patrick was sick so we had to communicate with him through call/text throughout the weekend).

Accomplishments that we're proud of

It actually works! (uploading PDFs, AI is able to generate quality information, and back end is connected to the front end)

What we learned

How to persist through many errors and git conflicts.
How to coordinate the back-end functionality to front end, with each team member working on a different portion (back-end generating, back to front end integration, and front end display).
How to fix GIT MERGE CONFLICTS.
Elisabeth learned how to use React for the first time.
Patrick learned how to use flask and tesseract for the first time.

What's next for Slack Off

We would have loved to have used MongoDB to store previous log entries so that users can track their past study content. This way, by recognizing their existing knowledge base, users can receive content recommendations based on their individual progress and preferences. Additionally, we hope to integrate the OCR API so that our AI can read text from images within PDFs, and speed up the time it takes for the AI to generate, as it takes quite a few minutes at the moment, depending on how much text it needs to parse through. We hope to deploy this to the public soon!