<![CDATA[VirCle is a mobile-based social media application that was developed as a medium for communication and interaction between fellow users. Application user data is stored in the Firebase database, without any additional security mechanisms. Even though Firebase provides security services when data is transmitted as well as access management mechanisms, because Firebase is a back-end database service, where the accessing application and the data are on a separate layer, the stored VirCle user data can still be accessed, either by other external applications through programming interfaces, or users with specific roles in the system such as administrators. On the other hand, the existence of stored user data is very important and must always be kept confidential by VirCle application service developers and service managers. Therefore, this research aims to implement the Advanced Encryption Standard (AES) algorithm for encrypting VirCle application user data so that it has an additional layer of security. The encryption design uses the AES-128bit algorithm considering the speed of the encryption and decryption process which is relatively faster than other symmetric cryptographic algorithms. The AES implementation in this study uses the Kotlin language by utilizing the BouncyCastle 128-bit AES library and CBC operating mode and PKCS5 padding to assist the encryption and decryption process. The evaluation results show that the developed VirCle application can encrypt user data (plaintext) into coded data (ciphertext) which is then stored in the Firebase database, then successfully reads and decrypts the stored ciphertext data from the database for reuse. Evaluation of application performance from the speed aspect shows that the encryption and decryption process time for user data is quite fast with an average total time of 1.35 milliseconds for the encryption process, and 1.53 milliseconds for the decryption process. Apart from that, evaluating the Avalanche Effect (AE) from the results of the AES-128 bit encryption algorithm used, an average percentage of 50.36% was obtained, this shows that the level of data randomness is considered good.]]>
