Grow your business with secure and scalable generative AI services from AWS and SAP
INTRODUCTION KEY INFORMATION ARCHITECTURE CUSTOMER USE CASE SUMMARY BIBLIOGRAPHY AND RESOURCES

Grow your business with secure and scalable generative AI services from AWS and SAP

AWS AI SAP AWS BEDROCK SAP BTP
October 3, 2024•AWS•SAP•Amazon Bedrock•SAP BTP
Key services Amazon Bedrock•SAP BTP•Generative AI Hub•SAP AI Core
A problem to solve Generative AI integration with SAP to optimize business processes and improve operational efficiency.
Benefits for the customer
  • Simplified data protection,
  • Reduced risk of data loss,
  • Faster recovery,
  • Central management
  • Compliance and security,
  • Cost efficiency,
  • Flexibility in data management,
  • Seamless disaster recovery,
  • Private and secure backups
  • Better operations management with generative AI,
  • New conclusions and optimization of business processes,
  • Business process automation.

INTRODUCTION

Dynamic development of generative AI technology combined with the SAP Business Technology platform?

In today's dynamically developing business world, implementing modern technologies is crucial for long-term success. Many organizations that run mission-critical SAP workloads on Amazon Web Services (AWS) are looking for innovative solutions, especially as generative AI grows in popularity. To support this trend, AWS and SAP have expanded their strategic partnership by introducing generative AI capabilities through Amazon Bedrock models, now available in SAP's AI Hub.

This integration enables companies to build AI-powered applications on SAP Business Technology Platform (BTP) while maintaining system integrity.

KEY INFORMATION

For SAP HANA workloads, it offers the following features:

Amazon Bedrock, a fully managed service, enables enterprises to develop generative AI applications using base models (FM) from leading AI providers such as Anthropic, AI21 Labs and Amazon Titan. Key features include model personalization based on company data, task-specific tuning, and Retrieval Augmented Generation (RAG) to improve response accuracy. Bedrock also supports the creation of intelligent agents that automate business processes, while ensuring solid security and compliance with standards such as SOC and ISO.

AWS also offers AI infrastructure services, such as AWS Trainium and AWS Inferentia, that enable companies to efficiently train AI models and perform inference in the cloud at lower costs. Additionally, Amazon Q Developer, an AI-powered coding assistant, increases developer productivity by generating real-time code suggestions based on natural descriptions and previous code snippets. Together, these tools enable companies to take an innovative approach to operations and modernization, while maintaining the high security and reliability standards required for SAP workloads.

ARCHITECTURE

SAP has been an early adopter of generative AI technology, introducing Generative AI Hub on the SAP Business Technology Platform (BTP). The AI ​​Hub within SAP AI Core offers AI tools tailored to your needs, access to leading base models, and full control over your data to build AI-based applications. By integrating with Amazon Bedrock models such as Anthropic's Claude3 and Amazon Titan, SAP customers can improve their business processes using generative AI. This common reference architecture (JRA) allows SAP customers to adopt generative AI across a variety of use cases, such as RISE with SAP, intelligent scenario lifecycle management, and custom applications built on SAP BTP. This integration delivers new insights in key business areas such as finance and human resources, with plans to further develop these AI capabilities.

diagram

ELEMENTS

Amazon Bedrock enables customers to access a wide range of Large Language Models (LLMs) via API, including base models such as Amazon Titan and Claude Anthropic. Pre-trained on massive datasets, these models are versatile and capable of supporting a variety of use cases. Customers can use them as-is or customize them privately with their own data.

Generative AI Hub in SAP AI Core provides access to AI models and manages their lifecycle, acting as an interface to SAP applications running on SAP BTP. It provides LLM models like those from Amazon Bedrock while implementing strict content filtering, risk mitigation, and regulatory compliance. This ensures secure and scalable integration of generative AI into the SAP ecosystem. The Hub also connects to the vector capabilities of SAP HANA Cloud, which enables you to deliver more contextual and accurate answers by including business data in model results.

SAP HANA Cloud, with its multi-model database management system, supports intelligent data applications at scale. It includes a vector engine that supports Retrieval Augmented Generation (RAG) to improve the performance of language models by providing better context.

The Cloud Application Programming (CAP) model simplifies the development of cloud applications on SAP BTP. CAP offers a framework for data modeling and integration, accelerating development through a structured approach. In this JRA, CAP is used for application logic while SAP UI5 handles the UI.

Large language models often ignore user authorizations during queries, which poses an integration challenge. SAP BTP and SAP Identity Provisioning services solve this problem by managing the identity lifecycle both inside and outside SAP systems, ensuring secure and authorized interactions with models.
Using these components, customers can build scalable and reliable SAP applications powered by generative AI. This common reference architecture not only supports business process extensions within SAP environments, but also helps maintain the SAP-recommended clean core approach.

CUSTOMER USE CASE

A manufacturing company using SAP BTP integrates generative AI with Amazon Bedrock to automate recruiting processes. Thanks to generative models such as Amazon Titan, the system processes candidate applications, automatically creates rankings based on competencies and generates initial job offers. This solution speeds up recruitment, increases selection accuracy and improves candidate experience.

A retail company uses SAP and Amazon Bedrock to optimize its supply chain. The generative AI system analyzes sales and warehouse data, predicts product demand, and then suggests optimal order dates and delivery routes. The company reduces operating costs, avoids excess inventory and improves the time it takes to deliver products to customers.

Financial services company implements generative AI with Amazon Bedrock for customer service on SAP BTP. The AI ​​model analyzes the history of customer interactions and generates personalized financial product recommendations. Thanks to integration with SAP CRM, the system can automatically create offers tailored to individual customer needs, which increases engagement and sales conversion.

SUMMARY

Integrating Amazon Bedrock's generative AI capabilities with SAP Business Technology Platform (BTP) gives SAP customers access to powerful Large Language Models (LLM), enabling them to improve their business operations with AI-driven insights and automation. This common reference architecture enables organizations to modernize key processes, optimize operational efficiency and reduce costs. By leveraging advanced data management from SAP and advanced AI models from AWS, companies can solve complex challenges and improve customer experiences through tailored, innovative solutions. SAP and AWS will continue to evolve this integration, focusing on specific use cases to unlock additional value.

BIBLIOGRAPHY AND RESOURCES

Power your business with secure and scalable generative AI services from AWS and SAP

ARTICLES AND BLOGS

Automate and Simplify SAP HANA Backups with AWS Backup

AWS BACKUP SAP

Read more

Integration of SAP systems with AWS services

AWS SAP

Read more