Inspiration:
The inspiration behind the Sakhi was to create a platform that helps travelers practice speaking a foreign language in a realistic and interactive way. The goal was to provide users with a personalized AI language learning experience that simulates real-life situations and helps them improve their language skills.
What it does:
Our website allows users to enact real-life situations in any language, such as ordering food at a restaurant or checking into a hotel. The website also provides users with helpful hints in case they are not sure how to reply in the conversation.
How we built it:
The project was built using a variety of tools and technologies, including voice recognition software, machine learning algorithms, and natural language processing (NLP) tools. The main NLP pipeline was built using Google's Speech APIs and the AI back-end interacts and exchanges information with ChatGPT via its open-source APIs.
Challenges we ran into:
One of the biggest challenges the development team faced was creating a realistic and engaging user experience. The team had to come up with a way to simulate real-life conversations with ChatGPT, which required a lot of testing and iteration to get right. Additionally, the team had to ensure that the website was accessible and user-friendly for a diverse range of users so we had to focus majority of our time on the UI/UX.
What's next for Sakhi:
The team behind plans to continue improving the platform, incorporating new features that allow Sakhi to correct the user's pronunciation and vocabulary so that they can improve their language in the limited time they have.
GitHub repo:
Built With
- chatgpt
- css3
- gcp
- html
- javascript
- python
- react
Log in or sign up for Devpost to join the conversation.