Senior RTMP / Video Streaming Architect (IoT Cameras + Mobile App)
Project Overview We are white-labeling RTMP-enabled IP cameras from a Chinese OEM and need a senior-level streaming architect to finalize, validate, and document a scalable, multi-tenant RTMP architecture integrated with our own RTMP ingest server and mobile application (iOS/Android). The goal is to support thousands of cameras across many customers using a single RTMP ingest endpoint, while enforcing logical isolation by camera ID and account, not by URL. ⸻ Current Components • RTMP-enabled OEM cameras (configurable before shipment) • Central RTMP ingest server (cloud-hosted) • Backend API / database • iOS & Android mobile app • QR-code–based onboarding flow ⸻ Target Architecture • Single RTMP ingest URL (e.g. rtmp://ingest.ourdomain.com/live) • Unique stream keys per camera • Stream key derived from camera UUID / serial / token • Account isolation by camera ID • Not by RTMP endpoint • Server-side stream validation • Authenticate stream keys at ingest • Map camera → user → account in backend • Pre-shipment camera configuration • OEM loads RTMP URL • Camera generates or is assigned stream key / camera ID ⸻ End-User Flow (Must Be Supported) 1. Camera powers on 2. User connects camera to Wi-Fi 3. Camera displays QR code (camera ID / token) 4. User scans QR code in mobile app 5. App claims camera → user account 6. Live stream becomes available immediately ⸻ Scope of Work • Design and validate multi-tenant RTMP ingest architecture • Define: • RTMP URL structure • Stream key format and lifecycle • Camera authentication & authorization strategy • Evaluate: • Whether to use static vs rotating stream keys • Push vs pull RTMP models • Define best practices for: • Preventing stream hijacking • Handling reconnects, dropped streams, and retries • Scaling ingest and playback (HLS/WebRTC optional) • Specify OEM configuration requirements for camera firmware • Review QR-code onboarding flow for security and reliability • Provide: • Architecture diagram • Sequence diagrams • Written technical recommendations ⸻ Required Expertise • Deep hands-on experience with RTMP ingest pipelines • Experience with: • NGINX-RTMP / SRS / Wowza / custom RTMP servers • Camera or IoT video devices • Stream key authentication • Multi-tenant backend systems • Familiarity with: • RTMP → HLS/WebRTC pipelines • Mobile app camera onboarding • Bonus: • OEM camera firmware workflows • Chinese camera manufacturers (Tuya, HiSilicon, etc.) ⸻ Deliverables • Final RTMP ingest architecture • Stream key + camera ID strategy • OEM camera configuration spec • Onboarding flow documentation • Security & scalability recommendations ⸻ Timeline • 1–2 week engagement • Immediate start ⸻ Screening Questions (Required to Apply) 1. RTMP Architecture Explain how you would support thousands of cameras using a single RTMP ingest URL while ensuring each camera streams securely and is isolated per customer. 2. Stream Keys What is the role of RTMP stream keys, and how would you design them for a multi-tenant camera platform? (Include how you would prevent stream hijacking.) 3. Camera Onboarding A camera ships with an RTMP URL preloaded. How do you securely associate that camera with the correct end-user account at first launch? 4. Failure Handling How should the system handle dropped RTMP connections and reconnects without creating duplicate or orphaned streams? 5. Bonus (Optional) Describe your experience working with OEM or white-label IP cameras and any RTMP servers you’ve deployed at scale. Apply tot his job