Circa

Circa is a flashcard-based learning platform designed to enhance conceptual understanding by discouraging rote memorization.

The platform utilizes a Next.js frontend and a MongoDB NoSQL backend, capable of supporting thousands of users.

To evaluate a user's conceptual understanding, Circa quizzes users with items they're trying to learn, such as new words (e.g. prescience) or concepts (e.g. Gauss's Law). Users may choose to respond with text or speech, which is then processed using a HuggingFace Wav2Vec2 speech-to-text model. User responses are then embedded and semantically compared to the correct answers provided during setup to accurately assess understanding.

Users have the option to discuss each concept they're trying to learn with a personalized LLM, equipped with an understanding of each users' goal and learning style.

All of this processing is done on Circa's backend, which can be quickly accessed using a Python FastAPI server request.

The primary goal of Circa is to shift the focus of learning platforms, such as Quizlet, from mere memorization towards genuine understanding.