Ascolta

Welcome to Ascolta! This repository contains the backend code for the Ascolta website, which leverages AI to help people learn languages more effectively.

Features

Ascolta currently offers two key features:

1. Live chat with a chatbot: Users can practice their language skills with a chatbot that role plays a character with its own unique personality. The chatbot uses OpenAI's APIs for text generation, and can provide users with instant feedback on their language skills.

2. Listening and reading comprehension exercises: Users can practice their listening and reading skills through conversations between characters, which include audio, a transcript, and questions to test comprehension. Ascolta uses Azure's Cognitive Services SDK for Text-to-Speech generation and Azure's Storage SDK for storing generated audio and images.

Project Structure

This repository contains all the backend logic for the Ascolta website. The project is organized into the following directories:

• src: This directory contains all of the backend logic for Ascolta.
  • ai: This directory contains the logic for AI-generation functions.
  • audio: This directory contains the logic for audio-generation functions.
  • config: This directory contains configuration files for integrations, such as environment variables.
  • db: This directory contains the logic for database integration.
  • routes: This directory contains the endpoints that are exposed on the server.
  • scripts: This directory contains scripts that can be run directly on the server to avoid exposing them via HTTP.
  • storage: This directory contains the logic for storage integration.
  • types: This directory contains custom types used throughout the codebase.

Technology Stack

Ascolta is built using a variety of technologies and services, including:

• OpenAI API: Ascolta uses OpenAI's API for text generation, allowing the chatbot to simulate realistic conversations with users.
• Azure Cognitive Services SDK: Ascolta uses Azure's Cognitive Services SDK for Text-to-Speech generation, enabling the site to generate high-quality audio files for listening exercises.
• Azure Storage SDK: Ascolta uses Azure's Storage SDK for storing generated audio and images.
• Azure Cosmos DB (MongoDB): Ascolta uses Azure's Cosmos DB for persisting structured data.
• Node.js: Ascolta is built using Node.js, a popular JavaScript runtime for building scalable web applications.
• Express.js: Ascolta uses Express.js, a minimalist web framework for Node.js, to handle HTTP requests and responses.

License

This project is licensed under the GPLv3 License - see the LICENSE file for details.