Offline reliability
Downloaded tours have to work with no connection, images and media included. Some image-loading code quietly assumed a network was available, so offline tours behaved inconsistently. I tracked down where image and media loading still relied on the network or cache, and fixed it so downloaded tours work consistently offline.
Sign-in flow
Sign-in and sign-up had grown into one large, hard-to-follow flow, so even small changes were slow and risky. I restructured it around a single host flow with separate screens for entry, OTP verification, password sign-in and account creation, and made navigation and back-stack behavior more predictable. The flow is now easier to follow and safer to change.
Background work
Newer Android versions restrict what apps can do in the background, and parts of the app needed to catch up without breaking what already worked. I moved suitable jobs to WorkManager and made notification and state cleanup more reliable.
Purchases, crashes and analytics
Some problems spanned the app and several third-party services: crashes, purchase-flow issues, and inconsistent analytics or revenue tracking across Firebase, Amplitude, Branch and Facebook. I traced each report back through the system involved — the crash report, the specific SDK or the purchase flow itself — to find the real cause.
