SEEKING SENIOR SDET - Device Spoofing & Session Orchestration Layer (Platform-agnostic) I am hiring an experienced Java SDET / Mobile Automation Engineer to build the entire Device Spoofing & Session Orchestration Layer v1 for a high-scale mobile automation system. This milestone is platform-agnostic and must work for any app. The system must support rooted Android devices running Magisk + LSPosed, with Android Faker, ProxyDroid/SocksDroid, and a GPS spoofing app installed. Your responsibility is to deliver a production-ready module that controls full identity spoofing, network/GPS coherence, c l o n e-instance binding, Appium execution, and per-device session management. This is not simple UI automation — this is device-level identity rotation and deterministic session control, built in Java with Appium 2.x and ADB. You will also build a NAT Chaos Emulation Layer, which generates randomized daily schedules of 48–96 potential sessions per device (does not need to run all 96 yet), assigns random gaps, c l o n e s, fingerprints, proxies, and geos, ensures no identical (fingerprint, proxy) pair is reused in the same day, and outputs a queue of session requests. The milestone also requires Appium Grid Integration, including querying real devices from the Grid, mapping UDIDs to node URLs, generating capabilities dynamically (UDID, c l o n e package, activity), and a clean appiumdriverfactory that starts drivers reliably on real hardware (no mocks). You will implement a C l o n e <--> Identity Binding System, where once an account is created or assigned to c l o n e #X on device Y, that c l o n e is forever “home” for that identity. This system must store and return bindings using a persistent DB. A full Unified Data Model must be provided, including entities/tables for Device, App C l o n e, DeviceSession, DeviceState, FingerprintProfile, NetworkProfile, PlatformAccount (generic, not platform-specific), and C l o n e Binding. ADB utility wrappers must be included for shell execution, airplane mode toggling, GPS apply, proxy apply, app launch, app force-stop, app data clear, and log retrieval. You must also deliver a minimal SessionWork interface, with a sample implementation that simply opens the c l o n e d app, performs a few UI interactions, and exits. Everything must be fully documented in a README explaining how to run it, how devices are registered, how c l o n e s are added, how fingerprints rotate, how proxy/GPS works, how to run a test session, and how to interpret logs. Acceptance Criteria: (1) Three full sessions must run successfully on at least one real device end-to-end, each with a unique fingerprint, unique proxy, matching GPS location, correct c l o n e launch, working Appium driver, and clean session shutdown. (2) DB tables must populate correctly for DeviceSession, FingerprintProfile, NetworkProfile, and CloneBinding. (3) The NAT scheduler must generate a day schedule with the correct number of randomized session slots. (4) The Appium Grid integration must successfully run a sample SessionWork UI interaction using a real c l o n e app. This milestone includes everything listed above; no shortcuts or mocks.