Computer Vision & Live Streaming Expert Needed (WebRTC, FFMPEG, NVR/DVR, Django, React)

Project OverviewWe have an existing Computer Vision & Live Streaming Platform that integrates WebRTC, FFMPEG, NVR/DVR sources, Django, FastAPI, React, and more. However, we are facing several technical challenges that are preventing us from scaling and deploying across multiple use cases. We are seeking a highly experienced individual or team specializing in real-time video streaming, scalability, and fault tolerance to troubleshoot, optimize, and enhance the system for production readiness. Retry logic implemented but not functioning reliablyIntermittent stream failuresNeed a strong auto-recovery fault-tolerant solution2.Camera ScalabilityCurrent limit: ~8 camera streams per serverMust scale far beyond this while optimizing CPU/GPU/memory usageThe same setup behaves differently on different servers3. Configurable Multi-Use Case UINeed a single UI framework that can be configured for multiple CV use cases, such as:-Attendance tracking-Construction site monitoring-Fire detection-Theft / pickpocket analytics-Number plate recognition (ANPR)-Rather than writing new code per use case, we need a config-driven or drag-and-drop solution for:1-Dashboards2-Alerting3-Reporting4-Layouts4.Environment Consistency / DeploymentCode behaves differently across serversNeed standard deployment scripting or containerizationPrefer Docker/Kubernetes bolthires/CD approachWhat We Expect✔ Deep troubleshooting of reliability and retry logic✔ Scalability optimization (more cameras per server)✔ Environment standardization: Docker/K8s bolthires/CD✔ Modular configurable UI design✔ Documentation + Knowledge TransferIdeal Candidate-Proven experience with WebRTC + FFMPEG + NVR/DVR-Past work scaling video streaming systems in production-Strong in Python (Django / FastAPI) + ReactCloud DevOps (AWS, Azure, GCP)-Microservice architecture, Docker & KubernetesStrong problem-solving skills & ability to deliver production-quality solutionsDeliverables-Reliable live streaming with working retry logic-Scalable architecture supporting higher camera loads-Configurable UI framework for multiple use cases-Deployment documentation-Knowledge transfer sessionScreening Questions (Must Answer)1-Describe your experience using WebRTC + FFMPEG for live streaming from NVR/DVR sources.2-Have you scaled camera streaming beyond 8–10 streams per server? What approach did you use? 3-How do you ensure environment consistency across multiple deployment servers? 4-Share an example of a configurable UI/dashboard you built that supports multiple use cases without additional coding. 5-What techniques or tools do you use for real-time monitoring and fault tolerance in streaming systems? 6-How do you ensure security in live streaming and NVR/DVR integrations? Please include:1-Relevant project examples2-High-level approach to solve our challenges3-Estimated timeline & pricing model Apply tot his job