Retailers today increasingly deploy and maintain business applications at edge and within that grapple with availability, latency and bandwidth bottlenecks with applications like CCTV video platforms streaming high-resolution videos 24/7. This session explores edge computing architectures paired with scalable video platform design, showcasing how Kubernetes and open-source tools enable real-time video stream and processing at the edge. Learn from a real-world retail deployment managing 200,000+ cameras, achieving sub-second latencies and hybrid storage strategies. We dissect architectural trade-offs, security hardening, and lifecycle management for distributed edge-to-cloud video workflows.

Key Takeaways

• Edge Computing: Leveraging Kubernetes clusters for low-latency video processing, prioritizing localized compute and dynamic scaling.
• Video Platform Optimization: Reduce cloud dependency via edge filtering, open-source streaming pipelines.
• Retail-Specific Outcomes: Achieve real-time security alerts, bandwidth/cost reduction, and compliance with hybrid storage architectures.

Target Audience

• Edge Infrastructure Engineers: Building fault-tolerant, distributed systems for video workloads.
• Video Platform Developers: Optimizing streaming pipelines for latency-sensitive applications.
• Retail Technology Teams: Scaling CCTV analytics across stores, warehouses, and cloud backends.
• Kubernetes Architects: Managing stateful edge clusters in resource-constrained environments.