Divyanshu Joshi

Zero-Downtime IT/OT Infrastructure: The AI-Ready Network Foundation for Smart Factories

Abstract:

As factories become increasingly dependent on AI, edge computing, autonomous machines, digital twins, and real-time analytics, traditional definitions of network uptime are no longer sufficient. A plant may remain technically connected while still experiencing jitter, stale telemetry, delayed inference, synchronization failures, or degraded machine-to-machine communication that can disrupt production. This talk reframes zero downtime as continuity of the complete production decision loop and presents an architecture for AI-ready IT/OT infrastructure built around resilience, deterministic performance, security, observability, and bounded degradation. The proposed architecture integrates physical control systems, deterministic connectivity, edge intelligence, an industrial data fabric, enterprise and AI platforms, and a cross-layer control plane. The session explains why fault containment is as important as redundancy and outlines recovery strategies for failures across networking, edge compute, stateful processing, cloud platforms, applications, and security controls. It also distinguishes the timing, latency, and reliability requirements of hard real-time control, robotics, AI inference, operational analytics, and background workloads. Security is extended beyond users and devices to machines, workloads, AI models, and industrial processes through context- aware Zero Trust principles. The talk also traces the evolution from monitoring to self-healing infrastructure through telemetry correlation, dependency graphs, anomaly detection, causal reasoning, predictive analysis, controlled remediation, verification, digital twins, and governed AI agents. Together, these capabilities provide a foundation for safer, more resilient, and increasingly autonomous smart factory operations.

Brief Profile:

Mr. Divyanshu Joshi is a Leader of Solutions Engineering at Cisco Systems, where he leads AI-driven infrastructure programs supporting a $500 million Amazon business portfolio. He has directed engineering initiatives that improved Amazon’s AI data center deployment efficiency by 30% using Cisco Silicon One, Acacia optics, and Cisco 8000 Series systems. He has also co-developed solutions with Amazon engineering teams, authored Customer Requirements Documents (CRDs), and helped accelerate Cisco-led AI initiatives by 40%. Previously, he managed more than $120 million in annual engagements and delivered large-scale infrastructure solutions, including network infrastructure supporting Panasonic’s 6- million-square-foot AI-ready smart factory. His expertise spans end-to-end IT/OT architecture, data center networking, cloud and edge infrastructure, and enterprise security using technologies such as Nexus 9000, UCS X-Series, SD-WAN, and Zero Trust frameworks for global manufacturers. His leadership has been recognized through Cisco Manager Excellence and nomination to Cisco’s Director Development Program. He also hosts Cisco’s “Embracing Change” Leadership Podcast, where senior leaders share perspectives on organizational transformation and effective change leadership. In addition, he has authored research presented at ICSSAS, IEEE AIC, and ICCNT, served as a peer reviewer for IEEE AIC, and judged innovation forums including the Claro Awards, HackRice 15, and HackTX 25. Throughout his career, he has been recognized for influential communication, cross- functional leadership, analytical problem-solving, and his ability to align diverse stakeholders while delivering scalable solutions with measurable business impact.