NTT DATA Business Solutions improves SAP code translation and generation with Azure

NTT DATA Business Solutions wanted to make it faster and easier for its internal and customers’ developers to integrate SAP with other key systems and processes. This included streamlining the time-consuming manual translation of specifications.

The company built a software as a service (SaaS) assistant using the Azure platform and Azure OpenAI Service generative AI capabilities. It uses Azure Cosmos DB, Azure Kubernetes Service (AKS), and others to manage and optimize its Virtual SAP Consultant.

By automating code translation and generation with Azure OpenAI Service, NTT DATA Business Solutions has significantly reduced time and costs for developers who use the Virtual SAP Consultant while improving accuracy and document standardization.

Companies in all industries sift through numbers to find the slightest advantage over their competition, and they all agree—data is king. NTT DATA Business Solutions helps a diverse array of customers harness the power of data to drive business transformation in the digital age. With more than 30 offices and 16,000 employees around the world, NTT DATA Business Solutions offers many innovation-enabling services, including application management, automation implementation, and hybrid cloud transformation. Its parent company, NTT, is one of Japan’s largest publicly traded companies. “Regardless of the baseline technology, we create innovative solutions for the pain points facing our customers,” says Hande Odabasioglu, Marketing Director at NTT DATA Business Solutions. The multifaceted technology company serves as a global leader in SAP consulting, providing end-to-end services and solutions to help customers unlock the software’s full potential.

The challenge of integrating SAP solutions

When implemented effectively, SAP can help businesses optimize their operations and increase efficiency by automating processes and organizing data. On the other hand, integrating SAP with other key systems and processes can be challenging sometimes. “When you want to connect third-party tools and applications—like creating and serving APIs with a Microsoft .NET solution—SAP can feel like a closed box,” says Gorkem Aslan, Software Engineering Team Lead at NTT DATA Business Solutions. “We help customers make the most of their SAP implementations.”

SAP users frequently use custom code to manage critical business processes, but SAP platform customization can be labor intensive. It typically involves functional consultants specifying and tailoring solutions before engineers implement them. To streamline these time-consuming tasks, NTT’s growing team of SAP experts based in Istanbul, Türkiye, set out to develop a tool to accelerate time to value and improve customer outcomes with the power of generative AI. “We cannot change the speed of the market. We need to change our ways of working by adapting to the AI phase of the world,” says Arda Calli, Head of Innovation at NTT DATA Business Solutions. The team determined that its AI-driven solution needed both a cloud platform and an integrated suite of technology features to automate code translation and generation while providing in-depth program analysis. NTT DATA Business Solutions selected Azure and Azure OpenAI Service with Azure AI Foundry to provide the bedrock of its groundbreaking Virtual SAP Consultant digital assistant.

Streamlining SAP activities with Azure OpenAI Service

NTT DATA Business Solutions had a clear understanding of the approach it would take. “We needed to deploy an application with the cloud because we wanted to build a SaaS [software as a service] solution,” says Aslan. “We containerized the application in Azure so that if you want to deploy a non-SAP solution, like a .NET or React JavaScript app for web development, you can still continue with a separate back-end application when deployed on Azure.” NTT chose this strategy to enable its Virtual SAP Consultant to integrate with other systems and applications that customers may already have in place.

We containerized the application in Azure so that if you want to deploy a non-SAP solution, like a .NET or React JavaScript app for web development, you can still continue with a separate back-end application when deployed on Azure.

Gorkem Aslan, Software Engineering Team Lead, NTT DATA Business Solutions

The team also wanted to reduce the time it takes for SAP users to manually translate functional specifications (FS) into technical specifications (TS). The Virtual SAP Consultant uses the Azure OpenAI Service GPT-4o model to automate this time-consuming process. Using the assistant simplifies FS document creation and writes TS documents, helping developers and others explain and document their processes in a specific format. “We use Azure translation systems and many other Azure AI services with our custom rack solutions and normalization solutions,” says Aslan.

The Virtual SAP Consultant also generates Advanced Business Application Programming (ABAP) code and provides seamless ABAP program analysis to further assist SAP consultants and facilitate better business decision-making. “You can enter inquiries about SAP through strategically developed input fields and the assistant will generate intelligent answers using the latest GPT-series model,” notes Aslan. “We’re doing prompt engineering with our own experiences so we can guide our customers and give them the best answers and best possible solutions.” This intuitive interface enables developers with varying experience levels to produce their intended SAP output without additional support.

You can enter inquiries about SAP through strategically developed input fields and the assistant will generate intelligent answers using the latest GPT-series model.

Gorkem Aslan, Software Engineering Team Lead, NTT DATA Business Solutions

Maximizing the value of data

To store these questions and carefully apply the data they generate to inform future query responses, NTT DATA Business Solutions uses Azure Cosmos DB. “We use Azure Cosmos DB for getting requests and responses in a backlog and storing session data such as when someone signs in, what the request was, and when it was sent,” explains Aslan. “If there is a failed request, we need to understand why it failed and how to prevent a similar request from failing in the future. Having requests, responses, and session information available in Azure Cosmos DB allows us to gain that understanding.” The company also uses Azure Key Vault to protect this valuable information. “We are using Key Vault to keep our back-end application environment variables private.”

Data protection is key, but so is organization. NTT has implemented a queue system with Azure Service Bus to help ensure the Virtual SAP Consultant can respond in an orderly fashion to all the queries it receives. The ever-expanding amount of data generated from these queries is invaluable for NTT as it continues to enhance the insights its Virtual SAP Consultant can provide. To store and leverage this massive amount of data, the company adopted a microservices-based approach, using Azure Kubernetes Service (AKS) to manage and scale their microservices. “We can create, release, and deploy to different environments easily,” says Aslan. “We monitor and manage all of our microservice applications easily within AKS.”

More about this diagram

NTT DATA Business Solutions also has a large number of developers to manage, and it uses Microsoft Entra ID to determine each employee’s permissions when building and maintaining its Virtual SAP Consultant. “When I want to give a developer a role, create a group for team members working on specific applications, or grant access to customer viewers, I manage those roles with Microsoft Entra ID,” says Aslan.” This helps NTT bolster the security of its innovative SAP assistant while using GitHub Enterprise helps streamline its workflows. “We use GitHub with repository management in the deployment phase,” he continues. “That workflow runs in the background while we build applications.”

Increasing efficiency and expanding capabilities

Developers within NTT DATA Business Solutions and its customers now use the Virtual SAP Consultant, available on Azure Marketplace, to overcome previous custom code–related barriers to upgrading or migrating to new SAP offerings. The company’s developers have found that the assistant automatically converts FS documents into TS documents and generates ABAP code with outputs of up to 80% compliance with SAP best practices. This generates an estimated 70% reduction in development time and 50-70% reduction in costs compared to an average development report process. These promising indicators are a sign of innovative capabilities to come for the solution that was just released in late 2024 and continues to expand daily. As Calli highlights, “The Virtual SAP Consultant is more of an ongoing journey than a finished product. We have a roadmap with questions to answer in the first phase so we can better understand the needs of our customers going forward.”

With this roadmap in hand, NTT DATA Business Solutions plans to continuously expand and optimize its AI assistant—and further establish itself as an industry leader. “We’ll add new features in the near future, and we’re offering our customers access to the OpenAI Academy program to train their developers to build their own solutions,” says Calli. As data lights the path forward, NTT DATA Business Solutions and its Virtual SAP Consultant are setting a new standard for innovation.

The Virtual SAP Consultant is more of an ongoing journey than a finished product. We have a roadmap with questions to answer in the first phase so we can better understand the needs of our customers going forward.

Arda Calli, Head of Innovation, NTT DATA Business Solutions