I need a high-performance Python application that runs natively on Windows and brute-derives 12-word BIP-39 mnemonic phrases, checks the resulting Bitcoin and Ethereum addresses, and immediately flags only the ones holding large balances. Speed is critical: the program must fully utilise my CPU cores, minimise external calls, and keep the scan loop as tight as possible so I can cover significant keyspace quickly. When an address with a substantial balance is detected the private key (or full seed) should be transferred straight into my own wallet file or through an automated import routine I can trigger locally—no cloud hand-offs or third-party storage. A real-time console (or lightweight GUI if that’s easier in Python) showing current scan rate, addresses tested, and any “hits” will help me monitor progress. Deliverables • Windows-ready executable plus clean Python source • README with setup, dependencies, and instructions to add or switch API endpoints for balance lookups • A simple configuration file where I set the minimum balance threshold that defines “large” Acceptance criteria: the tool must consistently exceed one million derivations per minute on a modern multi-core CPU, correctly identify known funded test addresses I provide, and import each winning key into my designated wallet without manual copy-paste. Feel free to leverage libraries such as bip-utils, pycryptodome, requests/urllib3, and asyncio or multiprocessing for parallelism as long as the final package remains self-contained and easy to run on Windows.