Good morning and good afternoon everybody. My name is Ana Yata, and today I'll be walking you through how AI transformed the incident management. In supply chain operations. So from whole reactive process to a proactive process via automated prevention system, a lot of this thinking builds on both academic work and my professional experience including my time at Amazon where we dealt largely with a complexity of large scale operations and the importance of minimizing the disruptions, especially the supply chains. The critical challenge is basically like cast cutting a supply chain disruptions, supply chains are usually a fragile ecosystems, a single failure delayed shipment. Or a warehouse breakdown can create a ripple effects across the whole network. At Amazon, I saw this firsthand. One node in the system such as a regional fulfillment center could trigger stockouts and missed delivery promises. The downstream traditional approaches usually tend to focus on fixing the problems after they occur, but by that point, the customer trust and revenue are already at risk. This is a challenge we are trying to solve. Next is AI revolution from reactive approach to a proactive approach. As I said earlier, here is a fundamental shift instead of waiting for a breakdown and then scrambling in the AI enables us to predict and prevent. At Amazon, for example, we used a predictive analytics to forecast ASIN level product overrides before they disrupted the sales, essentially intervening before the customers felt the pain. The same principle applies here, the you need to automate the detection, gain the full visibility, and then proactively resolve the issues. Next is a comprehensive data stream analysis. To do this, AI systems must ingest multiple streams of data. For example, supplier metrics, the transportation timelines, inventory flows, whether even geopolitical risks. No single analyst can connect all these dots fast enough, but machine learning can. At Amazon, we often layer clickstream. Supply chain and vendor data to get the whole 360 degree view of operations. This holistic perspective is what drives early detection and smart interventions. Next is machine learning pattern recognition. Machine learning builds a baseline of what's normal looks like in your operations. Once that's established anomaly stand out immediately, whether it's a spike in transit delays or unusual supply lead times that you have. At Amazon anally detection models helped us to identify when a particular product class was deviating from its usual demand supply chain pattern leading us to adjust it before escalating into a larger problem. The key is moving from a hand site to a foresight. Next is ink. Iot enabled real time warehouse visibility. Warehouses are the heartbeat of supply chains. The iot sensors make them smarter by tracking everything. Like temperature, the humidity and the vibration, and lot with respect to equipment health as well. This data creates a digital nervous system. For example, at Amazon, we tracked a conveyor performance and environmental conditions to spot the bottlenecks before the disrupted The order flow iot paid with AI enables us to prevent further actions rather than a costly downtime. Next is computer vision and predictive maintenance. Computer vision is like giving eyes to machines. It detects wear and tear patterns in equipment, misaligned packages, or early failure indicators. Coupled with predictive maintenance algorithms, organizations can schedule fixes before breakdowns can occur during mind time at Amazon. The predictive maintenance on sortation equipment saved us countless hours of downtime and directly impacted the delivery sles. Next is reinforcement learning for response optimization. When incidents do occur, reinforcement learning ensures the right response. These systems learn from historical responses, what worked, what didn't, and optimize the routing of alerts to the right teams. Think of it as a continuously learning playbook. At Amazon, we often refined alert routing logic to reduce the noise for ops teams ensuring critical incidents got immediate attention while also avoiding the alert fatigue. Next is advanced predictive escalation systems. These systems don't just like flag issues, they predict escalation. For example, models can anticipate when an incident breach might when you know, like when an incident might breach. SLA thresholds and trigger automated migration protocols. Faster than even like human intervention. So this is crucial in high velocity businesses like Amazon where even few minutes of delay in addressing the fulfillment incidents can cascade into thousands of missed deliveries. Next comes to the federated learning for collaborative intelligence. Supply chains aren't also, they're not isolated, so they are as they are ecosystems of partners. Federated learning lets companies share threat intelligence without exposing sensitive data. Imagine multiple partners contributing insights on disruptions, collectively strengthening the network. In finance, we call this as consortium data in supply chain. It's basically a collaborative resilience. Next two comes to the blockchain integration for audit ID trails blockchain ensures also the trust and compliance. It provides immutable audit trails and also enables the faster and smarter contracts for automated compliance reporting. At Amazon, while we didn't use a blockchain directly for incident management, we heavily invested in immutable logging for Audible. So similar in principle to the blockchain, ensuring the integrity across the large scale systems. Next is implementation of strategy framework. Of course, transformation requires a structured approach. We need to start with assessment around the pilots to validate then scale gradually. At Amazon, we followed similar playbooks when deploying the new data science tools. Begin small through the impact and then expand systematically. Continuous optimization is also always like vital as new risks and data patterns emerge. Next is operational benefits and impact. The benefits are tangible. Like they reduced meantime to resolution. Fewer escalations higher C higher service reliability at Amazon, every percentage point improvement in uptime or delivery reliability translates into usually like millions of dollars saved and the customer trust gains. The same benefits extend here. Proactive systems save time, money, and reputation. Finally, the future of supply chain operations. The benefits are tangible here as well. So the future is about like residence and then adaptability. AI will transform supply chains from firefighting to foresight driven. Orchestration. Organizations that invest now will thrive as networks get more global and complex. My experience at Amazon showed me that scale and complexity demand nothing less than AI power solutions. Next is, thank you for your time. So if you have any questions feel free to drop an email to me. So my email is krishna c dot ya@gmail.com. Thank you.