Our models — including facial recognition, speech-to-text, and screen-sharing segmentation — are trained exclusively on publicly available data from YouTube, which is copyright-free. We have not used, nor do we plan to use, customer data to train our models. These models are developed in-house, reflecting our commitment to copyright compliance and data security. In addition to our proprietary models, we also utilize ChatGPT for enhanced functionality.
​
We prioritize data security and video data safety across all our services. Your videos are securely stored on AWS servers in California, protected by the AES-256 encryption algorithm. Access to a video for editing is facilitated through a temporary encrypted link, unique to each user and valid for two hours, ensuring your content remains secure and private. Furthermore, videos are automatically deleted seven days post-upload for free users, with an option for immediate deletion available through the workstation. All the deleted projects are permanent from both the user's workspace and our backend and will not be able to recover.