Data Security
Data Encryption
All user-related data will be encrypted within the database, ensuring that even if unauthorized access were gained, the data remains unreadable without the appropriate decryption key.
Salt Vault and User- Specific Encryption Keys
Upon the first login, a unique salt vault is generated for each user, securely storing the encryption key needed to decrypt user-related data. This system, similar to the end-to-end encryption used by platforms like WhatsApp and Telegram, ensures that decryption can only occur on the user’s device. This means that neither we nor any other party can access the user’s data without the encryption key tied to the user’s device.
Data Protection in Case of Breach
Even if the encrypted data were somehow made public, no one would be able to decode it without the user’s unique encryption key. If a user loses their password.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Optional: Military-Grade Infrastructure Obfuscation
For enhanced security, we can implement military-grade infrastructure obfuscation. This involves using multiple proxies and a sharded data model, where no single server contains the entirety of the user’s information. This method prevents potential attackers from gaining access to all data from any single server. Additionally, we can further strengthen security by adding a decryption key field at login, requiring users to provide their decryption key alongside their credentials.
Best Practices for Infrastructure Security
Our infrastructure follows industry-leading security practices, including secure access control mechanisms, continuous monitoring, and compliance with relevant data protection regulations.
IP Address Privacy
While client-side security measures prevent us from masking the IP address, we ensure that no IP addresses are recorded or stored in our database.
Handling LLM (Language Models) and AI Services:
Any data sent to GPT-based models is anonymized to protect user identity. For instance, if a user inputs “Hi doctor, my name is Diego,” GPT receives “Hi doctor, my name is User717171.” The original name is restored on the client side, maintaining privacy throughout the process.