Good morning everyone, and thank you for joining me today. As global supply chain becomes more complex and interconnected, the risks we face from fraud to equipment failures are not just increasing, they are compounding. In fact, recent studies shows that a single disruption can cascade and impact nearly half of downstream operations. But what if your supply chain could see this? These risks coming and fix them before they cause damage. Today I will walk you through a cutting edge, a powered framework for anomaly detection and self failing in supply chains. We will explore how advanced techniques like graph neural networks, transformers, and blockchain back. Smart contracts are D driving a real-time decisions, automating responses, and saving millions both in dollars and downtime. Let's dive into the future of resilient and intelligent supply chains, the growing complexity of supply chains. Let's begin by understanding the landscape. Fortune 500 companies now work with over 10,000 suppliers across more than 10 TH one 30 countries. This incredible web of connections introduces vulnerabilities at every level. As a result, we have seen a two 17% increase in risk exposure, and when a critical node fails, 43% of downstream operations are impacted. How about the financial consequences? An average of one $84 million in annual losses per organization. This clearly shows a traditional reactive approaches no longer surface the real time event processing framework. To address this, we need real time responsiveness and with the help of Apache Flink and Kafka streams, these platforms allow us to process massive volumes of events with millisecond latencies. For example, Flink handles up to 500 events per second with the latency under 8.3 millisecond and Kafka mayonnaise 23.4 milli million events per minute. All this is orchestrated under Kubernetes with K native, ensuring 99.995% availability and seamless scaling for event driven systems, advanced anomaly detection with ai. Now, how do we find disruptions before they escalate? GNS are key supply chains naturally form graphs and GNS help detect anomalies within these complex networks with 92.7% accuracy and 68.3% reduction in false positives. They fair outperform traditional methods. Transformer based models known for handling time series data identified 76.2% of anomalies at least 14 days before conventional methods. With fine tuning, we achieve 83.7% accuracy using just 200 labeled anomalies. Explainable A with powerful models, explainability is crucial. We use sharp values to highlight which features influence predictions with the top three features, often contributing 64.7% of the outcome, reducing investigation time by nearly half. Counterfactual explanations. Let operators test what if scenarios Improving intervention success by 32.6% and attention visualizations. Highlight suspicious nodes in the transaction graph, achieving 86.3%, source identification in under four minutes. The self-healing capabilities, detection is only part of the puzzle. What happens after an anomaly is detected. Our framework supports full self feeling, kept full self feeling using Apache airflow. We orchestrate workflows that dynamically adapt to disruption scenarios. The meantime to resolution has dropped from 27.3 hours to just 4.8 hours. 82.4% improvement. We detect, analyze, remediate, and continuously learn. It is a full lifecycle of intelligent response. The automated inventory rebalancing a helps prevent problems, not just fix them. Through dynamic rebalancing, the system reduces stockouts by 64.7% and cuts excess inventory by 37.8%, saving $3.4 million annually for a mid-size company. All while maintaining a service level above 98.5%, the smart contract validation via blockchain. Blockchain technology ensures trust in a decentralized ecosystem. Using Hyperledger Fabric, we validate transactions securely and eliminate 97.4% of disputes with the settlement times improving by 89.3%. Smart contracts enforce compliance automatically removing 73 manual checks per transaction and achieving a hundred percent accuracy across 2 37 transaction types. Human AI collaboration, even with automation, humans are still critical. We use chart ops bots integrated with Slack and Microsoft teams to alert users quickly. Cutting awareness time from 73 minutes to just 9.3 minutes. Our knowledge base helps operators to solve normal issues, 41.7% faster, and through continuous learning loops, the system evolves with each incident, reducing false positives by nearly 9% every quarter. Everything is good. However, the performance and results metrics, let's wrap up with impact. 87% reduction in undetected fraud. 62% decrease in disruption duration, 43% improvement in inventory metrics, 91% of anomalous detected within five minutes. Most organizations saw Voes under 10 months, and clearly nearly 90% exceeded their expectations. These are not marginal gains. They are transformative. In summary, modern supply chain demand more than just visibility that require intelligence and autonomy. With AI at the core, organizations can detect anomalous in minutes, reduce disruptions significantly, and optimize operations with confidence. The future of supply chain resilience is not just automated, is not just only automated, but explainable, secure and human centered. With this, I close, my presentation. Thank you all for your attention. I look forward to see your questions and thoughts.