The proposed framework describes a comprehensive approach like Learning with errors, Ring Learning with errors to secure and process sensitive data created in the age of artificial intelligence. Starting with data readiness to ensure data is ready, the process encrypts the data using strong algorithms and transforms it into encrypted string data for secure data transfer to the cloud platforms. Shortest Vector Problem and Closest Vector Problem helps in increasing the complexity of the algorithm to enhance security. Encrypted string data maintains data confidentiality and is only accessible to authorized parties with the proper secret key. Various types of data are accepted, except for visual datasets such as images or video. Encryption converts the data to lattice points. A public key is included to hide the information. Adding noise during encryption protects against eavesdropping. During decryption, noise is separated from the secret key’s tiling by the secret key itself. It is important to use a good basis to decryption in order to prevent an intruder from decrypting the wrong ciphertext. The goal of this framework is to ensure data safety and confidentiality through Post-Quantum Cryptography the entire data lifecycle and the results are pretty satisfactory with encryption time of 0.40547 ms and decryption time of 0.71531 ms for public key size of 800 bytes, secret key 1632 bytes and ciphertext 768 bytes. Testing and security measures are implemented to mitigate vulnerabilities.