Introduction:
In my previous blog, I shared an example of how the SAP HANA Cloud vector engine and generative AI hub were utilized in an SAP Partner solution to achieve AI-driven innovation in the energy procurement domain. This was just the beginning of a series, where we explore how different partners are harnessing SAP’s advanced technologies to drive innovation. Today, we continue this journey by looking into one of Deloitte’s pioneering AI use cases, developed in collaboration with SAP’s BTP Gen AI Lighthouse Program. Deloitte has implemented the GenAI Supply Chain Planning Assistant solution – demonstrating the impressive capabilities of the SAP HANA vector engine within SAP’s technology stack.
Use Case: Gen AI Supply Chain Planning Assistant
Problem Statement:
Supply chains are dynamically complex, businesses must mitigate disruptions to operational efficiency by effectively managing exceptions and minimizing inventory discrepancies. Inventory shortages and the inability to fill customer orders can lead to financial losses. The root cause is often the lack of timely access to essential information, which causes delays and suboptimal decision-making. This leaves supply chain planners struggling to identify issues and make informed decisions swiftly. Exception handling becomes particularly difficult when unexpected problems arise, such as stock-outs despite having sufficient inventory aligned with replenishment plans. Organizations cannot afford to overlook these supply chain issues and their underlying causes. Without comprehensive visibility into the end-to-end supply chain, the entire company is affected.
Solution:
Deloitte describes the Gen AI Supply Chain Planning Assistant as a tool that enables supply chain planners to efficiently manage exceptions, optimize inventory, and identify the root causes of disruptions. By leveraging AI-informed decision-making, the assistant aims to help save time and money while enhancing overall productivity.
This tool provides a macro-level dashboard that gives an overview of the entire network, pinpointing exceptions that need immediate attention. It highlights issues such as unexpected stock-outs despite having sufficient inventory levels aligned with replenishment plans
The assistant responds in near real-time, presenting insights and visualizations to help resolve the issue. For example, it might reveal that while forecasts and safety stock levels were correctly established and purchase orders generated on time, a failure occurred because the supplier did not ship according to the agreed timelines. This rapid, AI-driven feedback allows planners to address and resolve exceptions promptly, maintaining the smooth flow of the supply chain and minimizing potential disruptions.
This use case depends on the SAP HANA Cloud vector engine to enable key functionalities like:
Embedding Business Knowledge into Vectors: To analyze and consolidate data from unstructured sources, like vendor communications that impact delivery or emails between sales and demand planners. Additionally, the tool can extract insights and assist in root cause analysis by reading vendor agreements or contract PDF documents.Embedding Technical Knowledge into Vectors: This improves the accuracy and performance of business query responses within the AI Assistant chatbot by embedding successful and accurateInterestingly, for new business queries, the chatbot first triggers a similarity search in the SAP HANA Cloud vector engine (before invoking the LLM) to retrieve context from the previously generated or executed highly similar SQL queries. The reference to previously generated SQL queries maintains the consistency & accuracy of the response.
Furthermore, the solution utilizes generative AI hub to access Azure OpenAI’s GPT foundation models to facilitate prompt engineering, perform RAG workflow, predict inventory levels, and identify potential disruptions.
Reference Architecture:
Value add of SAP HANA Cloud vector engine in this use case:
Unified Database: Allows consolidation of data from diverse sources into a single database (vector data can be now stored alongside other structured business data), enabling the development of innovative inventory management applications.Vectorization of supply chain vendor communications (emails), contracts (PDFs), and demand forecasts (Excel) enhances inventory management by capturing context, extracting critical terms, and facilitating accurate demand predictions.Optimized Query Handling:Stores SQL queries for successful prompts.Retrieves stored SQLs for identical prompts in future queries.Uses similarity search to find and adjust the nearest SQL query as per the given prompt.Reduced Latency and Cost: By minimizing interactions with the LLM model, it reduces latency and operational costs.Improved User Experience:Provides enhanced search capabilities.Offers personalized recommendations, making the application more valuable and user-friendly.
Conclusion:
In exploring Deloitte’s Gen AI use case, we see how the SAP HANA Cloud vector engine plays a critical role in advancing supply chain planning by providing a robust foundation for AI-driven insights and decision-making. This use case also highlights the power of SAP’s BTP to solve complex business challenges.
Be sure to look for my next blog post, where I will highlight another exciting use case from the SAP BTP Gen AI Lighthouse Program. We will dive into how another partner leverages AI and advanced technologies to drive innovation and efficiency.
Note: This blog is a collaborative effort between SAP HANA Cloud Product Management and Deloitte, highlighting the innovative use case enabled by the SAP HANA Cloud vector engine. For those interested in implementation details of this solution, Navjot Sharma (nasharma@deloitte.com) and Nipun Pratap Singh (nipsingh@deloitte.com) from Deloitte are available as the primary contacts.
Next Steps:
Get a hands-on with SAP Developer’s Code Jam with SAP HANA Cloud vector engine & generative AI hubIntelligent Data Apps Are the Future: Read through this e-book to learn whyListen to the podcast on Journey to Intelligent Data Applications: Modernize with SAP HANA CloudGet a sneak peak into the use cases our partners are building through the SAP BTP Lighthouse Program
Introduction:In my previous blog, I shared an example of how the SAP HANA Cloud vector engine and generative AI hub were utilized in an SAP Partner solution to achieve AI-driven innovation in the energy procurement domain. This was just the beginning of a series, where we explore how different partners are harnessing SAP’s advanced technologies to drive innovation. Today, we continue this journey by looking into one of Deloitte’s pioneering AI use cases, developed in collaboration with SAP’s BTP Gen AI Lighthouse Program. Deloitte has implemented the GenAI Supply Chain Planning Assistant solution – demonstrating the impressive capabilities of the SAP HANA vector engine within SAP’s technology stack.Use Case: Gen AI Supply Chain Planning AssistantProblem Statement:Supply chains are dynamically complex, businesses must mitigate disruptions to operational efficiency by effectively managing exceptions and minimizing inventory discrepancies. Inventory shortages and the inability to fill customer orders can lead to financial losses. The root cause is often the lack of timely access to essential information, which causes delays and suboptimal decision-making. This leaves supply chain planners struggling to identify issues and make informed decisions swiftly. Exception handling becomes particularly difficult when unexpected problems arise, such as stock-outs despite having sufficient inventory aligned with replenishment plans. Organizations cannot afford to overlook these supply chain issues and their underlying causes. Without comprehensive visibility into the end-to-end supply chain, the entire company is affected.Solution:Deloitte describes the Gen AI Supply Chain Planning Assistant as a tool that enables supply chain planners to efficiently manage exceptions, optimize inventory, and identify the root causes of disruptions. By leveraging AI-informed decision-making, the assistant aims to help save time and money while enhancing overall productivity.This tool provides a macro-level dashboard that gives an overview of the entire network, pinpointing exceptions that need immediate attention. It highlights issues such as unexpected stock-outs despite having sufficient inventory levels aligned with replenishment plansThe assistant responds in near real-time, presenting insights and visualizations to help resolve the issue. For example, it might reveal that while forecasts and safety stock levels were correctly established and purchase orders generated on time, a failure occurred because the supplier did not ship according to the agreed timelines. This rapid, AI-driven feedback allows planners to address and resolve exceptions promptly, maintaining the smooth flow of the supply chain and minimizing potential disruptions.This use case depends on the SAP HANA Cloud vector engine to enable key functionalities like:Embedding Business Knowledge into Vectors: To analyze and consolidate data from unstructured sources, like vendor communications that impact delivery or emails between sales and demand planners. Additionally, the tool can extract insights and assist in root cause analysis by reading vendor agreements or contract PDF documents.Embedding Technical Knowledge into Vectors: This improves the accuracy and performance of business query responses within the AI Assistant chatbot by embedding successful and accurateInterestingly, for new business queries, the chatbot first triggers a similarity search in the SAP HANA Cloud vector engine (before invoking the LLM) to retrieve context from the previously generated or executed highly similar SQL queries. The reference to previously generated SQL queries maintains the consistency & accuracy of the response.Furthermore, the solution utilizes generative AI hub to access Azure OpenAI’s GPT foundation models to facilitate prompt engineering, perform RAG workflow, predict inventory levels, and identify potential disruptions.Reference Architecture: Value add of SAP HANA Cloud vector engine in this use case:Unified Database: Allows consolidation of data from diverse sources into a single database (vector data can be now stored alongside other structured business data), enabling the development of innovative inventory management applications.Vectorization of supply chain vendor communications (emails), contracts (PDFs), and demand forecasts (Excel) enhances inventory management by capturing context, extracting critical terms, and facilitating accurate demand predictions.Optimized Query Handling:Stores SQL queries for successful prompts.Retrieves stored SQLs for identical prompts in future queries.Uses similarity search to find and adjust the nearest SQL query as per the given prompt.Reduced Latency and Cost: By minimizing interactions with the LLM model, it reduces latency and operational costs.Improved User Experience:Provides enhanced search capabilities.Offers personalized recommendations, making the application more valuable and user-friendly.Conclusion:In exploring Deloitte’s Gen AI use case, we see how the SAP HANA Cloud vector engine plays a critical role in advancing supply chain planning by providing a robust foundation for AI-driven insights and decision-making. This use case also highlights the power of SAP’s BTP to solve complex business challenges.Be sure to look for my next blog post, where I will highlight another exciting use case from the SAP BTP Gen AI Lighthouse Program. We will dive into how another partner leverages AI and advanced technologies to drive innovation and efficiency.Note: This blog is a collaborative effort between SAP HANA Cloud Product Management and Deloitte, highlighting the innovative use case enabled by the SAP HANA Cloud vector engine. For those interested in implementation details of this solution, Navjot Sharma (nasharma@deloitte.com) and Nipun Pratap Singh (nipsingh@deloitte.com) from Deloitte are available as the primary contacts.Next Steps:Get a hands-on with SAP Developer’s Code Jam with SAP HANA Cloud vector engine & generative AI hubIntelligent Data Apps Are the Future: Read through this e-book to learn whyListen to the podcast on Journey to Intelligent Data Applications: Modernize with SAP HANA CloudGet a sneak peak into the use cases our partners are building through the SAP BTP Lighthouse Program Read More Technology Blogs by SAP articles
#SAP
#SAPTechnologyblog
