Hey. Hi everyone. Good morning, good evening. My name is Kumar. I'm having 12 years of experience in S-A-P-E-R-P solutions In this conference we will discuss how generative VA is revolutionizing business intelligence. In cloud ERP ecosystems. In this presentation we will mainly focus on how generative artificial intelligence can evolutionize predictive analytics capabilities within ERP cloud systems. We'll examine how advanced machine learning models, particularly generative adverse networks, integrate with existing ERP infrastructures to transform forecasting accuracy across multiple business functions. And through the technical foundations and implementation strategies and cross industry case studies, we will demonstrate the tangible business value. While addressing critical challenges in data quality, system architecture, and model maintenance you will gain a strategic roadmap for leveraging generative AI to achieve competitive advantage and through enhanced operational efficiency and data driven decision making. So as an example, like Simon users generative VA in its team center, ERP system to simulate supply chain disruptions and generative optimal recovery plans. Improving resilience, so the next one, what we are looking at is evaluation and market trends. So we are we are differentiating the market trends by three different ones. So unprecedented growth. So the global AI in ERP market is expecting rapid expansion with projections indicating a compound annual growth rate of for 29.7%. Through the 2013, so the Microsoft Dynamics 365 as an example 365, dynamics 365 a integrates with ERP system contributing a projected by 30% CS year by 2030. So widespread adoption, the second market trend these widespread adoption, approximately 65% of large enterprises now incorporated some of some form of EA capability within their ER ecosystems. A significant increase from previous implementation rates. A, as an example, Walmart uses a ERP for inventory management and reducing the stockouts by 30%. And the next one, what we are looking at is industry leaders. So the accelerated adoption is particularly evident in manufacturing, retail and health healthcare sector. So where complex operational requirements that create a compelling use cases for predictive capabilities. So as an example, like Johnson employs EA in S-A-P-E-R-P to optimize pharmaceutical production schedules. So the next one, what we are looking at is the next one we discuss about, the next section, what we will discuss about is technical capabilities and business value. So here we have three different capabilities here. The first one is enhance decision making, organization implementing AI enhance. DRP systems reports some substantial improvements in decision making process with 78% of early adopters documenting enhanced operational outcomes through more accurate forecasting and predictive maintenance. As an example, Coca-Cola uses here and S-A-P-E-R-P to forecast regional demand and reduced waste by 20%. The next one is efficiency improvements. So company's reports an average reduction of 35% in manual data processing requirements on following successful deployment creating compelling financial justification for implementation. As an example Unilever automated invoice processing in its ERP cutting work cutting manual work by 40%. And so the next one is deployment models. So cloud-based implementation has become a dominant paradigm offering the scalability and computational resources necessary for advanced remodeling. Training and interference. So Netflix leverages S-A-P-E-R-P Cloud with AI for real time content licensing, cost analysis. Through that they are reducing a lot of cost which is getting which is getting implemented as part of the cost analysis. Okay, the next one, what we are looking at is architectural framework for AI and AI and ERP integration. So here we have a different layers approach, a file layer approach where we can integrate AI and ERP together. So the first one is se security layer where we will ensure data production, and compliance when we are integrating, with AI to ERP. And the next one is a technology layer where we provide the computing inter infrastructure which enables which enables the integration of AI and ERP and the next one where we manages the information flow and the storage of data using data layer. And the next one, how we handle the business logic and processing using the application layer. And the last one is a presentation layer where the user interface and experience is, delivered. So through, through these five layer approach, successful implementation follow this structured approach. Incorporating five key architecture layers of this framework enables organizations to maintain, clarity regarding system boundaries integration points and information flows while implementing advanced AI capabilities. As part of this security layer as an example, Pfizer ERP uses a blockchain for HIPA compliant data encryption in clinical trials. That is one of the, one of the biggest layer where everyone needs to ensure everything is secured. The data compliance data protection is one of the major major drive, data driven point when we are integrating with the A with ERP systems. So next one, what we are looking at is a cloud infrastructure on data management. How we, how the infrastructure. Infrastructure, so the next one, what we are looking at is a cloud infrastructure on data management. So in this section we will discuss how the infrastructure and data management can be handled. What are the benefits of cloud? And also how the integration works between the data and the data governance. So the first one, what we are looking at is cloud benefits. Some organizations implementing AI solutions experience substantial benefits from cloud deployment models within infrastructure, scalability, representing a primary. Advantage Modern Implementations leveraging the infrastructure as code approaches to ensure consistency and repeatability. As an example, Airbnb uses Global Cloud ca tools to scale ERP resources during the peak booking seasons. So the next one, what we are looking at is data governance. Successful implementations employ canon data models that normalize termin knowledge and formats across organizational boundaries. Organizations must establish clear policies regarding data ownership, on quality standards and use permissions. Proctor and Gamble. Procter and Gamble employs a canonical data model in HCRP to standardize global supply chain terminology. So as an integration approach and data pipeline automation maintains some data freshness with the real time and near real time data integration, enabling more responsive predictive models. Hybrid approaches optimize both data governance and analytical performance. Next one. What we are looking at is supply chain optimization through ai. So because of this optimization through ai, 23% average improvement compared to traditional time series approaches and also three plus successful implementation, combining ERP data with external factors, so which results substantial reductions while maintaining service levels. Organizations implementing year driven forecasting models experience significant improvements in supply chain operations, particularly when addressing demand volatility from market disruptions. The technical implementation typically follows a phased approach, beginning with data integration, followed by model deployment under culminating in interactive dashboards for scenario evolution. So as in supply chain example, Tesla combine CRP data with weather APIs to optimize battery supply chains, reducing delays by 25%. That is a huge reduction in delays for the Tesla. So the next one, what we are looking at is financial planning and production optimization. So as when we are implementing, mission learning into ERP. So we are looking at two, the three, two different techniques here. That is one is financial forecasting organization enhance some financial forecasting accuracy by implementing mission learning algorithms that identify complex relations between organizational metrics and financial outcomes. So JP Morgan uses mission learning in Oracle ERP to predict cash flow trends within 90% of accuracy. The next one is production scheduling companies implementing machine learning algorithms for production scheduling. Experiences average reductions of 18% in production lead times compared to traditional approaches. So Toyota Reinforcement Learning System in S-A-P-E-R-P cut the production lead times by 18%. So the reinforcement learning advanced implementations supply employee reinforcement learning techniques that continuously improve the scheduling policies based on observed outcomes gradually adopting to changing production constraints. The next one, what we are going to discuss about is how the manufacturing sector's implementation will impact the usage of machine learning in the ERP sector. So here we defined four different sections. The first one is the requirement gathering where where we will define business objects and the technical requirements. For the sector. And also the next one is data preparation, where we will validate data quality and integration diverse sources. As part of the development we will create and try and predict to algorithms so that those algorithms will be deployed at the time of, usage. And the fourth one is integration testing. Verify the system's compatibility on the performance as part of your integration testing. So an automotive parts manufacturers implementing and yay enhanced forecasting system that delivered remarkable operational implement some. This solution incorporated both supervised and unsupervised learning techniques to identify complex patterns in production. Variability. The traditional statistical methods had to fail to detect. So overall when we are implementing in the manufacturing sector, when we implement both European and machine learning, we will see lot of a lot of implementation cases. So as an example, BMWA enhanced ERP reduction, reduced product production variability by 30% using GI gans. Next one. The next one, what we are discussing about is how the retail and healthcare works. When we implement EA in the into E-E-E-R-P systems in ea. So why we are using for example, the retail demand sensing, in retail dimensions, say multi-channel forecasting with external variables. JARAs, Oracle ERP integrates social media sentiment analysis to adjust the inventory weekly based on the social media. They, they performed more more sales, based on the sentiment how the social media works. And the next one, what we are looking at is compliance integration, whereas HIPAA complement data handled protocols. In the healthcare system in the healthcare system, myo Clinic uses A-P-R-P-V-T-A to predict a patient admissions reducing wait times by 15%. And also the next one is markdown reduction, improved allocation on timely replenishment. When we use the ei, that's what it happens. Like the, it'll reduce, it'll improve allocation on timely replacement. The next one is healthcare resource planning, optimize staffing and capacity utilization. So when we integrate with the a multi-channel retailer deployed a comprehensive productive analytics solution that incorporates social media sentiment analysis under competitor pricing data, and regional economic indicators to generate more accurate sales forecast. Similarly, a regional healthcare provider implemented predictive analysis analytics to optimize the staffing levels under source utilization while maintaining the strict compliance requirements. So the, in the retail and healthcare sector we will get more benefits when we integrate ERP with the ai. The next one, what we are looking at is implementation results across industries. Here we'll look at multiple examples how how the ERP plus yay produced the results in a way. It enhances the overall industry. For example, automated manufacturing where they implemented SA PS four HANA as an ERP system using GaN GN based forecasting, where it results, where it result 38 point 37.88% of forecast accuracy improvement and 4.7 million annual inventory sales. This was what happened for, with the fold. Ford. Ford, SAP S-A-P-E-R-P solution. And the next one, what we are looking at is multi-channel retail. For example, let's take imagine as example here. So they used Oracle retail as an ERP solution, and used the time series with external variables. So which resulted in substantial markdown reduction and improved the allocation percentages for Amazon. The next one, what we are looking at is in healthcare, where they use the epic systems as an ERP solution. Where where used resource optimization models as an AI technology, so as in results it increases the it optimized the staffing and improved the capacity utilization. What we are looking at is data quality and preparation challenges. And here again, here we defined we differentiated into four to four different approaches. The first one is initial validation where we verify the data against predefined quality thresholds. And the next one is standardization. Transform to consistent format across sources. As an example, Nestle ERP cleanses supply supplier data. Using AI and improved PP work, we see by 35% when they, as part of the standardization. So the next one is enrichment. So here we have to add derived features and contextual information, which results in the final verification confirms the readiness of model in ingestion. So organization implementing Yay enhanced ERP systems must address several critical data challenges, including inconsistency across sources incompleteness of historical records and semantic ambiguity. These adopting semantic data preparation methodologies experience significantly higher implementation success rates compared to those making ad hoc approaches. Okay. In the next section, what we are looking at is integration architecture considerations. So here we look at successful integration requires, developing locally coupled frameworks that isolate AI components from underlying ERP systems enabling independent evolution while maintaining functional integrity. The technological architecture are generally incorporates some specialized in middleware components that abstract the complexity of underlying ERP systems providing standardized interfaces for AI services. Some. So here, when we compare with the, between different layers, like middleware layer, event driven architecture, EAPI, component integration patterns, error handling. So when we are comparing between implementation complexity and the business impact the business impact is more the implementation complexity is there, but still the business impact because of the implementation is huge. That's where we need to consider implementing AI plus ERP. The next one, what we are talking about is model management and governance. So as part of this we need to look for five different section. So the first one, what we do is create and prime models with appropriate validation as part of the development. Once the development is completed, verify the performance against and performance ag business requirements. Sir. This is as part of the validation. Once the business requirements were verified, then we will implement the same in production with monitoring capabilities. And then we'll monitor to track the performance and trigger alerts whenever they require. And then finally, we retire replace outdated models with improved versions that will results in that will give them more more higher results. So organizations implementing robust model, governance processes experience substantially lower operational incident while maintaining higher prediction accuracy over extended periods. The technical implementation typically incorporates some automated monitoring capabilities that track model performance against established metrics triggering alerts when degradation exceeds predetermined thresholds. What we are looking at is future direction. So here in this section, we'll discuss how multi multimodal AI and edge computing, will result in different sectors, so multimodal AI integration, advanced systems that process diverse data or data types simultaneously, including structured ERP data with unstructured content like text. Images, audio on video to create more comprehensive predictive models with enhanced accuracy on the expandability. As an example, GI Healthcare combines MRI images with the RP data to predict equipment maintenance. The next one is edge computing architecture, the distributed intelligence that processes data locally at operational endpoints including manufacturing, facilities retail locations and logistics centers reducing the latency while enhancing the resilience and data. For geographically dispersed operations. Shell, this one used by Shell, shell deploys a GA in oil rigs to process sensor data locally reducing the latency by 60%. The next one, what we are looking at is model optimization. Is advanced techniques including. Quantization pruning and the knowledge of distillation that reduce computational requirements while preserving critical predictive capabilities for deployment across diverse computing environments. So this is a, this is the future. How it is going to be look like, so the next one, what we are looking at is the roadmap. So roadmap and conclusion. So as part of this we will look at the phased implementation strategy, assessment and organizational alignment. So as part of the strategy assessment evaluate organizational readiness. Some and identify high value use cases that align with business objectives. Establish clear success metrics on governance frameworks for implementation journey. And the next one, what we use is ask the, we need to follow the phased approach. Follow a structured approach with proof of concept validation, scale development, and continuous improvement. This enables a progressive capability building while demonstrating value incrementally. So the next one is organizational alignment. Established a dedicated center of excellence, combining technical specialists. Business domain experts under change management professionals to ensure successful adoption and ongoing optimization. The integration of generative AI with ERP cloud system represents a significant advancement in enterprise technology capabilities. Organizations that effectively harness these capabilities gain substantial advantages to improve forecasting accuracy and enhanced resource allocation and more agile response soon. Responses to market fluctuations. Hey, thank you. Thank you everyone for attending this conference. Please reach out to me on if we have any questions regarding implementing ERP and AI artificial intelligence e in ERP systems. So that will result that will give huge benefits to all the industry sector. So thank you.