FlashLearn

Inspiration

Learning new technologies is hard. Browsing through complex or poorly written documentation is frustrating for beginners and experienced devs alike. We wanted to accelerate the learning process by auto-summarizing and organizing any technical concepts the user searches for.

What it does

With FlashLearn, users can eliminate hours from the process of learning a new technology and making their own flashcards. Our platform uses intelligently-generated flashcards that quickly aggregate the most relevant technical information on any topic the user searches for. Since we pull data from some of the most reputable sources on the internet, our information is comprehensive, reliable, and accurate.

How we built it

We used Modal, Convex, and TypeScript for the front-end and back-end. We used OpenAI's API and GPT-3 for creating the intelligent flashcards. We also used DevDocs to combine information about multiple APIs and programming frameworks/languages.

Challenges we ran into

We also wanted to scrape technical info from FreeCodeCamp's sitemap by parsing each URL in the XML file and formatting it as HTML, and we tried a couple of different methods, ranging from using a ready-made scraper called Octoparse to learning Beautiful Soup. Unfortunately, none of the use cases worked for us. We spent hours looking through documentation for Beautiful Soup and different web scraping apps - and we could have saved a lot of time if a tool like FlashLearn had already existed!

We also tried to use the Material UI framework with TypeScript but the code in the docs did not work the same way in our code.

Accomplishments that we're proud of and what we learned

We learnt most the technologies we used over the weekend: like Convex, Modal, and Material UI, AI, LLMs, OpenAI. We're still new to AI, an we struggled with planning out the project and integrating the different technologies.

What's next for FlashLearn

We want to fine-tune our project and add more data sources to pull from - like GeekForGeeks and FreeCodeCamp. We also want to add multimedia support and accessibility features to make sure our tools is accessible to all kinds of learners.

Built With

• convex
• langchain
• material-ui
• modal
• openai
• python
• react
• typescript