Senior Python/AsyncIO Engineer: Fix "Thundering Herd" & Non-Blocking Architecture

We have an existing Python automation script (using nodriver and asyncio) that interacts with a high-traffic web portal. During peak traffic hours (a classic "Thundering Herd" scenario), the target server becomes unresponsive. Our current fetch request hangs for 15+ seconds, causing a cascade of TypeError: Failed to fetch errors. Your Mission (Fixed Price: $300 - $400): 1. Thundering Herd Mitigation: Implement a strictly non-blocking architecture. You must use asyncio.wait_for (or equivalent JS AbortController inside the browser) with a strict timeout and an immediate retry loop. The bot must never hang, even if the server drops the connection. 2. Form Automation: Automate a multi-step form (dynamic fields, file upload, final click) using nodriver. I already have the exact HTML source code of this form. Requirements: Expert in Python, AsyncIO, and nodriver. Experience with high-frequency, low-latency systems (trading, real-time data aggregation).