Real-Time Streaming Automation Build

Замовник: AI | Опубліковано: 22.10.2025
Бюджет: 120 $

I’m putting together a production-grade, real-time automation system that will sit between seven adult live-streaming platforms—Chaturbate, Stripchat, BongaCams, CamSoda, CAM4, SkyPrivate, and JasmineLive—and our on-prem Linode server (Ubuntu/Docker). From day one the solution must support all of these platforms and treat “capture live chat and tip events” as the top-priority feature. Here is the core flow I need implemented: • Each platform’s chat and tipping events are scraped or intercepted, normalized to JSON, then pushed through a secure WebSocket tunnel to the backend. • The backend distributes that data to two main consumers: 1) an OBS/vMix overlay service (via OBS WebSocket API) for real-time on-screen effects, and 2) an AI reply service that crafts chat responses using OpenAI or another LLM. • Everything is containerized with Docker and exposed through both WebSocket and REST endpoints for internal tools. • A React + Node.js admin dashboard lets me view sessions, toggle automation, and review structured logs/history. The immediate milestone focuses on these deliverables: 1. Stable Node.js service that logs and forwards live chat & tip data from all seven sites in under 500 ms latency. 2. Modular WebSocket gateway with JWT or similar auth, fully dockerized for Ubuntu 22.04. 3. Admin dashboard skeleton (login, stream list, real-time log view). Nice-to-haves such as expanded browser automation, advanced overlay choreography, and deeper AI chat logic can follow once the fundamentals are in place. Tech stack expectations: Node.js (latest LTS), Puppeteer or Playwright for browser control, WebSocket & REST, React 18, Docker Compose, Ubuntu, secure coding best practices, and Git-based CI/CD. If you’ve built high-throughput WebSocket pipelines, automated browsers at scale, or integrated OBS before, I’d love to see your approach and a rough timeline.