Custom Software & Mobile Development

Detailed Technical Deep Dive: Custom Software & Android Development

Summary: This document outlines Jishnu Mahanta's architectural approach to constructing native Android utility applications, encrypting client-side databases, and engineering robust offline-first synchronization loops.

1. SQLite Database Encryption (SQLCipher & Android Keystore)

In native mobile development, client-side data protection is critical. Storing sensitive customer profiles, financial balances, or local scan logs in cleartext SQLite databases invites reverse-engineering exploits. I harden local databases by applying SQLCipher (AES-256 database encryption) and securing keys using the hardware-backed Android Keystore.

Below is the Kotlin pattern I implement using SQLCipher to initialize secure, encrypted databases inside native Android systems:

import android.content.Context
import androidx.room.Room
import androidx.room.RoomDatabase
import net.sqlcipher.database.SQLiteDatabase
import net.sqlcipher.database.SupportFactory

fun getEncryptedDatabase(context: Context, dbKey: ByteArray): MyDatabase {
    // Generate a SupportFactory with the raw byte key retrieved from the Keystore envelope
    const val DB_NAME = "secure_local_ledger.db"
    val factory = SupportFactory(dbKey)
    
    // Load SQLCipher libraries into memory
    SQLiteDatabase.loadLibs(context)
    
    return Room.databaseBuilder(
        context.applicationContext,
        MyDatabase::class.java,
        DB_NAME
    )
    .openHelperFactory(factory) // Enforce SQLCipher database encryption
    .build()
}

2. Offline-First Architecture & WorkManager Synchronization

Field agents operating mobile apps in remote regions of Assam frequently drop connectivity. To prevent user flow interruption or data loss, I deploy an offline-first local database design combined with transactional sync queues managed via Android's WorkManager.

The Offline-First Data Lifecycle:

1. User Action: The agent captures KYC data and GPS coordinates.
2. Local Commit: The data is instantly written to the local encrypted SQLite DB using Coroutines. The UI updates instantly.
3. Queue Registration: A sync request is logged inside the Room DB with a status of PENDING.
4. WorkManager Trigger: WorkManager starts a background sync worker. The task is configured with strict constraints (e.g. network connected, battery not low).
5. Auto-Retry: If the upload fails due to network drops, WorkManager schedules a retry using exponential backoff.

import android.content.Context
import androidx.work.*
import java.util.concurrent.TimeUnit

fun scheduleDatabaseSync(context: Context) {
    // Define constraints: Sync only when internet is connected
    val constraints = Constraints.Builder()
        .setRequiredNetworkType(NetworkType.CONNECTED)
        .build()

    // Create a one-time work request with exponential backoff retries
    val syncRequest = OneTimeWorkRequestBuilder<DatabaseSyncWorker>()
        .setConstraints(constraints)
        .setBackoffCriteria(
            BackoffPolicy.EXPONENTIAL,
            OneTimeWorkRequest.MIN_BACKOFF_MILLIS,
            TimeUnit.MILLISECONDS
        )
        .build()

    // Enqueue the work as unique, preventing duplicate sync queues
    WorkManager.getInstance(context).enqueueUniqueWork(
        "unique_db_sync_job",
        ExistingWorkPolicy.REPLACE,
        syncRequest
    )
}

3. Native Hardware Integration (CameraX & GPS)

To capture verifiable field records (such as solar panel rooftops locations), apps must interface directly with native hardware sensors:

• CameraX API: I configure camera inputs to automatically control exposure, crop photos to target aspect ratios, and optimize compression to keep files small (sub-500KB) while preserving legibility.
• Geofenced GPS Mapping: I hook into Android Location Services, applying location filters to verify that GPS coordinates represent physical site coordinates within ±5 meters before unlocking forms.