Inspiration
Acknowledging the workload of teachers, we set out to create a tool to ease their tasks and enhance the learning process. We recognized occasional inaccuracies in classroom discussions that could lead to student confusion. To rectify these situations real-time and ensure accuracy in information delivery, we created our tool. Additionally, it assists teachers in responding to student questions promptly. Our tool can also be utilized as a fact-checker in debates to foster civil and accurate discussions.
What It Does
AITA serves as a teaching and learning assistant. It assists teachers by detecting and correcting inaccuracies in their speech real-time, ensuring the delivery of precise information. It also handles real-time student inquiries, providing immediate answers. Besides, AITA works as a fact-checker in debates, pointing out inaccuracies to maintain factual integrity.
How We Built It
We used Vite and React for the front end and Flask for the back end. OpenAI API, Hume, and Speechly were used for information verification, correction, and real-time speech-to-text processing. For fact-checking, we used the WikiCheck API.
Challenges We Ran Into
Our main challenges included gathering data for model training and debugging API requests. As first-time users of many APIs, identifying and rectifying issues proved difficult.
Accomplishments That We're Proud Of
We successfully trained a neural network for the first time, managed to chunk text for ChatGPT, and implemented a fact-checking feature that provides additional resources to teachers.
What We Learned
As novice hackers, we gained experience using Flask and several APIs for the first time. This project enhanced our research and debugging skills when encountering new technologies.
What's Next For AITA
We aim to incorporate user authentication and data saving between sessions, enabling users to track their progress over time. Furthermore, we aspire to deploy advanced sentiment analysis to provide users with in-depth feedback on their presentation effectiveness.
Built With
- flask
- hume
- javascript
- openai
- python
- react
- speechly
- vite
- wikicheck
Log in or sign up for Devpost to join the conversation.