Add domain and subdomain handling in dataset preparation and admin routes; enhance data submission with new fields

Refactor audio processing to implement fade-in, end trim, and fade-out effects for improved sound quality

Update server-side audio trimming logic to adjust duration from 150ms to 200ms for improved accuracy

Update audio trimming logic to adjust duration and sample calculations for improved accuracy

Enhance PCM data processing by implementing trimming at both start and end, improving audio quality and consistency

Enhance audio processing by implementing server-side trimming and updating metadata handling for PCM recordings

Refactor audio saving process to use PCM data with wave module, enhancing compatibility and flexibility in audio file handling

Remove unused function generate_verified_parquets from prepare_dataset.py to streamline code and improve maintainability

Adjust sync interval to 3 minutes; refactor SiriWave initialization and resize handling for improved performance

Enhance login functionality and UI; add email visibility option for OAuth2, adjust admin CSS for responsive design, and streamline audio metadata handling in storage

Refactor metadata handling in audio recording; streamline local file operations and remove unnecessary metadata storage calls

Enhance metadata handling and validation in recording storage; add transcription ID management and adjust scheduler timing

Update transcriptions table to include user_id and language fields; modify uploaded_at to use TIMESTAMP WITHOUT TIME ZONE. Adjust generate_verified_parquets function to use SYNC_INTERVAL for scheduling.

Refactor audio processing to remove Hugging Face upload functionality and add verified metadata handling for local storage; implement daily parquet generation for verified recordings.

Implement PostgreSQL metadata storage and create database tables for transcriptions and metadata

Refactor user session handling: switch to attribute access for Record objects, ensure session user is a dict, and improve stats handling in dataset preparation

Implement dataset synchronization with Hugging Face: add scheduler for automated uploads, update audio path structure, and enhance metadata handling for Supabase.

Add age group and accent fields to session form and update related logic

Refactor Supabase client initialization and Hugging Face upload logic to ensure proper authentication checks

Add saving state management to the recorder interface, disabling controls during save operations

Refactor dataset preparation to support language-specific audio storage and metadata management; streamline directory creation and improve error handling.

Refactor dataset handling by updating .gitignore, implementing stats and recordings management in parquet format, and modifying filename generation to use a shortened user ID prefix.

Add database migrations for recordings and profiles, implement user profile creation, and enhance metadata handling

Add initial project setup with Docker, environment configuration, and authentication middleware