TikTok Bot for Advanced Automation

Project Description:Hello,We are launching a specialized Social Media Marketing (SMM) production system and developing an advanced automation bot dedicated to TikTok. The entire architecture will be deployed on a Linux VPS. We already own the "raw material," which consists of thousands of high-trust, Aged TikTok Accounts. Our primary objective is to scale up and run over 500 concurrent sessions/requests simultaneously at ultra-fast speeds to meet the high demand of the Saudi and Gulf markets with premium quality.To maximize efficiency, eliminate excessive proxy bandwidth consumption, and drastically mitigate ban risks, we want to architect the bot using a "Hybrid Automation System" with two completely separate execution paths:Path 1 (Followers Service): This path must rely on Headless Browser Automation (e.g., Playwright or Puppeteer). It requires strict and precise browser fingerprint management (Canvas, WebGL, User-Agent, etc.) combined with rotating residential/mobile proxies assigned dynamically per session. This is critical to bypass TikTok’s strict anti-bot detection during follows.Path 2 (Likes, Saves, and Shares/Shares Actions): This path must completely bypass browser rendering. It should execute via ultra-lightweight, direct API calls (HTTP / POST Requests) with media, scripts, and asset loading entirely blocked. This path will integrate with a third-party signing API subscription (e.g., via RapidAPI or a custom solution to generate device signatures like X-Gorgon, X-Argus, and X-Ladon). The goal is to drop proxy consumption down to less than 2-5 MBs per 1,000 actions.Required Questions for Developers to Answer in Their Proposals:Is this hybrid design (Dual-path architecture) fully feasible, scalable, and stable on a Linux environment without critical technical bottlenecks?Is separating these actions the most secure and efficient approach to protect the aged accounts while reducing proxy bandwidth costs to near-zero?Is there any technical or security risk to the aged accounts if the same account alternates between Path 1 (via browser automation) and Path 2 (via direct HTTP POST requests with generated signatures)?What specific libraries, tech stack, and queuing mechanisms (e.g., Celery, Redis, RabbitMQ, Asyncio) do you propose to guarantee the absolute stability of 500+ concurrent processes?Thank you, and we look forward to reviewing your technical proposals.