Ettu - Your personal foreign language tutor

Imagine having access to a language tutor who speaks YOUR language - Has your interests, is available 24/7 and corrects your grammar. That's Ettu for you.

Comment

Inspiration

The inspiration for this project is the struggle of being an intermediate-level language learner, who knows that Duolingo can't take him to the next level. The only way to get there is immersion in the language, and outside of bothering your friend Fernando every week just to talk the little bit of the language you know, it's difficult to get that from 1000s of miles away.

What it does

What ettu does is allow persons to have access to their private language tutor, who is available to have intelligent conversations with them 24/7, about topics they're interested in.

How we built it

Building Ettu: A Breakdown of Key Steps

Building Ettu involved a combination of advanced technologies, thoughtful design, and meticulous planning to create an engaging and effective language learning experience. The process can be broken down into several key steps:

  1. Backend development: The Etk backend was built on Editor X with heavy support from Velo, a platform that enabled the seamless integration of various components required for the app's functionality.

  2. Custom elements: To enable access to the audio interface, a custom element called "foreign" was created. This element captures the user's voice, transcribes it using the built-in audio processor, and uploads it to the Chat GPT API.

  3. Speech recognition and synthesis: The app's code incorporates speech recognition and speech synthesis objects from the browser. These objects are used to recognize user input and speak back the translated text, creating an interactive and immersive experience.

  4. Repeater for messages: The Etk page employs a repeater that handles messages, clearing the conversation history for each new session. Custom buttons and animations were also developed to create a more engaging user interface.

  5. Velo integration: The app utilizes Velo to manage data, including loading the repeater, querying the database, and fetching member information such as user photos for a more personalized experience.

  6. Session storage: The app employs session storage to store chat history in memory, allowing for seamless conversation flow.

  7. Backend security: To ensure API keys remain secure, the logic for accessing the OpenAI API was moved to the backend through HTTP functions. This prevents unauthorized access to sensitive information by site visitors.

  8. Product research and user feedback: To refine the app and ensure it meets the needs of potential users, product interviews were conducted, and feedback was gathered. This allowed for the implementation of crucial features based on user preferences and requirements.

By combining these elements, Etk was successfully built as an innovative and immersive language learning tool that offers personalized, interest-based conversations with intelligent language tutors, available 24/7.

Built With

  • editorx
  • gpt-3.5
  • velo
  • wix-http-functions
Share this project:

Updates