Privacy and Security

At Snapclip, we take security and privacy very serious. Trust is the foundation of our relationship with our users, and we understand its importance. That’s why we implemented various security measures to ensure you can trust us from the very start.

We frequently receive questions from our users about privacy and security, so we felt it is important to address these valid concerns in detail.

Opt-in public

Recordings are never public by default. You will have to opt-in for sharing your recording with anonymous users. Team members will always have access to your recording because you explicitly allowed them to be part of your team.

Data Protection

A robust database security strategy is essential for protecting your data from unauthorized access, manipulation, or theft. We work with Supabase, a SOC2 Type 2 certified platform, to ensure your data is safe. We rely on Postgres Row Level Security to ensure your data cannot be accessed by anyone you have not explicitly permitted to do so.

• All customer data is encrypted at rest with AES-256 and in transit via TLS.
• Sensitive information like access tokens and keys are encrypted at the application level before they are stored in the database.

Screen Recording

When you start recording your screen, video chunks are securely transmitted from your browser via SSL-encrypted websockets and temporarily stored on our backend. When you finished recording, the video file is encoded into different formats and stored on our CDN. Each recording is assigned a unique, non-iterable ID, making it computationally infeasible to guess.

AI

For processing your recording, we utilize Google Cloud’s computational resources. Each analysis task is executed as a virtual process on physical NVIDIA A100 GPUs. Your recording is temporarily downloaded solely for analysis, with no data remaining persistent. After running the LLM and transcript models locally, the virtual environment is destroyed, leaving no trace behind.

We prioritize keeping your data secure by minimizing the use of third-party SaaS solutions. However, we make an exception for OpenAI’s GPT APIs, which we use to correct faulty transcripts, summarize recordings, and generate suitable titles. OpenAI ensures that its models are not trained using input from the API, as clearly stated on their Enterprise Privacy page.