Generative AI Hub Revealed at SAP TechEd 2023

At SAP TechEd 2023 SAP announced the company’s bold vision to be the leading business AI company by 2025. SAP brings Generative AI (GenAI) development to the forefront to improve the way companies operate.
Prior to the latest announcements on GenAI, SAP embarked on a collaborative journey with a selected set of partners within its ecosystem. This early engagement brought the partners closer to experience and build generative AI applications.
Partners were exposed to the new tools, streamlined interactions with large language models (LLMs), and the dynamic capabilities of SAP Business Technology Platform (SAP BTP)

Hack2Build – Special Edition on Generative AI

SAP brought out the first edition of Hack2Build to a few selected partners to pilot and build intelligent data applications leveraging generative AI hub.
Partners were challenged to create relevant AI use cases in the context of their customer’s real-world business processes by embedding generative AI into their solutions. 10 Partners were selected to pilot their use case with Generative AI Hub offered as part of AI Foundation. SAP’s unwavering support from the Partner Ecosystem Success organisation and the AI Product Engineering team enabled partners to rapidly construct enterprise applications, harnessing the potential of Generative AI in less than 7 days time. The first edition of Hack2Build on Generative AI was held during September 5 – 12, 2023.

Being the first in this series, the blog’s aim is to inspire readers to contemplate the extensive possibilities of Business AI use-cases as shown by the partners and encourage them to envision the application of generative AI technology in their own scenarios. For a start, the partners here demonstrated clearly what is possible with Large Language models (LLMs) harmonizing the SAP data with the Generative AI Hub. The inaugural edition of Hack2Build on GenAI, hosted by SAP in 2023, for the partners shows the impact that generative AI can have on current business processes and within specific industries.

Having piloted the early version of Generative AI Hub, the partners were invited to demonstrate their GenAI solution and share their experience with the larger SAP TechEd2023 audience. The innovations exhibited here not only brings immediate value to our customers but also paves the way for the future development by the partners on AI Foundation. Let us discover now the inspiring solutions, the partners built as a part of the Hack2Build initiative, which you can put into practise for your business.

Partner Innovations on Generative AI

Avvale Espana S.L.U

Contract document intelligence for OpenText with structured and unstructured data ​

Avvale Document Intelligence for Open Text enhances document management by leveraging Generative AI models. With the SAP S/4HANA system connected to Open Text, the GenAI solution addresses the document management complexities of handling multiple systems, external documents, and unstructured content, multiple languages, offering enhanced efficiency and user-friendliness. Processing documents and extracting semantic information stored in a vector and metadata database, the system features a conversational interface, allowing users to pose questions and receive contextually relevant answers. This approach supports the users on decision-making and supports intelligent and streamlined document management, providing organisations with a responsive solution for accessing contract information efficiently.

Avvale-Contract Document Intelligence for Open Text

Bosch Global Software Technologies

Collaborative Automatic Return Process for FMCG, Retail, and Manufacturing

This solution from Bosch tackles the substantial return volumes encountered in FMCG, Retail, and Manufacturing industries. It empowers customer service or warehouse clerks to streamline the return process by capturing images of returned items, and assess the return reason and the refund amount. By scanning return forms or sales invoices, the system extracts invoice data directly from SAP S/4HANA, performs quality and return policy condition validations, creates a return sales order and enables the return workflow automatically. GenAI plays a pivotal role in extracting details, performing validations, and generating content for emails to both approvers and customers, depending on the approval or rejection of the return. The use of SAP Document Information Extraction, SAP Cloud Application Programming Model (CAP), SAPUI5, and Prompt Engineering in Generative AI Hub ensures a swift formatting of emails and return notes, further enhancing the overall efficiency of the return workflow.

Collaborative automatic return process by Bosch

Capgemini

Intelligent Procurement Powered By GenAI

Capgemini’s solution is designed to facilitate prompt, intelligent, and effective sourcing decisions within the Supply Chain realm. The Generative AI hub based solution pioneers user-driven conversations, offering profound insights into critical parameters such as supplier performance, pricing dynamics, materials availability, lead time adherence and alternative material recommendations, thereby empowering users to make informed sourcing decisions. The quotations, audit files of suppliers from the source system and other data about the suppliers from third party files is converted to embeddings and stored as vectors. Prompt Engineering and Semantic search are used, and the ranked results are then passed to Open AI’s GPT-4 which delivers comprehensive responses to business users, ensuring timely and data-driven decision-making in procurement processes.

Capgemini- Intelligent procurement powered by GenAI

DXC Technology Services LLC

Tele-Parkinson Chat with GenAI

DXC’s Tele-Parkinson mobile application, designed for Parkinson’s disease patients, has been enhanced with the GenAI chat feature as part of the Hack2Build initiative. Serving as an extension to the existing application, this innovative addition supports doctors in making inquiries using both business and medical language, leveraging data collected from the Tele-Parkinson app. The solution is adept at understanding the context of application data, transforming clinicians’ ad-hoc questions into meaningful data insights and charts. SAP Datasphere and Generative AI Hub is used for Data Catalog embedding, knowledge grounding, Text-to-Query, and Text-to-Visualization. The chat application, leverages Large Language Models (LLMs), provides clinicians with descriptive and prescriptive data insights within seconds, ultimately enhancing the understanding of causal factors and treatment options for individual patients.

DXC Tele-Parkinson Chat with GenAI

Ernst & Young

Generative AI for Risk-mitigated autonomous contract creation

Ernst & Young’s solution enhances SAP Ariba Contracts with real-time risk-mitigated contract creation through Generative AI capabilities. The legacy contract clauses are extracted using Document extraction using OpenAI Embedding (text-embedding-ada-002), Facebook AI Similarity Search (FAISS), Text completion (text-davinci-003). The solution provides new contract clauses generation using pre-trained LLM model, an Object store for maintaining enterprise specific contract clauses and streamlined large volume contract creation. The user-friendly review and editing interface for approval, with semantic search capability and Responsible AI with harm data exclusion using curated prompts, ensuring ethical and compliant contract development processes.

Risk-mitigated autonomous contract creation by EY

Inflexion

ShiftFlex with SAP SuccessFactors and Generative AI

Adjudged as the 1st Runner-Up in Hack2Build

In addressing the limitations of existing tools and streamlining the often tedious work-shift planning process within departments, Inflexion’s ShiftFlex emerges as a desirable solution. ShiftFlex harnessed the power of Generative AI Hub and built a web-based application to manage work shifts. The inclusion of the chatbot, Flexer, enhances the user experience by providing conversation-based assistance for employee shift-related queries and schedule creation. Beyond this, ShiftFlex offers a visual shift planner, streamlined shift change requests, and approval workflows. Notably, the seamless integration with SAP SuccessFactors positions ShiftFlex as a comprehensive and efficient tool for modern work-shift management.

ShiftFlex by Inflexion

Infosys

Enhance Legal Compliance with Generative AI

Companies in industries dealing with hazardous materials, such as Life Sciences, Chemical, and Oil & Gas, must comply with legal requirements to accurately display warning and hazard text on material labels. Infosys has streamlined this process by utilizing the capabilities of SAP AI Core’s Generative AI Hub. The company extracts information from identified public, government, and regulatory sources through web harvesting, making it accessible to internal regulatory and compliance managers for confirmation and approval. The data is then processed through the GenAI algorithm to generate legally accurate translations for precautionary and hazardous statements based on specific codes. The integrated SAP Environment, Health, and Safety Management (EHS) system is employed to update phrase libraries for label master data, enhancing compliance, reducing costs, and improving reaction time to market demands.

Legal Compliance -Hazard Material Labels by Infosys

LTIMindtree

Smart ATP for Manufacturing – MTO and ETO Industries

Adjudged as the Winner in Hack2Build

LTIMindtree utilized SAP AI Core’s Generative AI Hub to enhance supply chain delivery through Smart Available to Promise (ATP). The focus was on addressing disruptions in the supply chain that can impact on-time delivery. Accurate delivery date predictions are vital to meet customer demands and avoid disruptions and increased costs. Generative AI played a crucial role in providing precise delivery date estimates by considering historical shipment data, real-time weather conditions, traffic patterns, production lead times, and other relevant factors. The Smart ATP solution, powered by Generative AI, is specifically designed to improve the reliability of ATP for engineer-to-order (ETO) and made-to-order (MTO) industries, ultimately enhancing trust and customer satisfaction.

LTIMindtree – Smart ATP for Manufacturing – MTO and ETO Industries

NTT Data Business Solutions AG

Simplified Corporate Sustainability Reporting with Generative AI

Adjudged as the 2nd Runner-Up in Hack2Build

Starting in 2024, the Corporate Sustainability Reporting Directive (CSRD) mandates approximately 50,000 companies, including non-EU entities with EU subsidiaries or listings, to adhere to European Sustainability Reporting Standards (ESRS). These standards involve disclosing over 80 key performance indicators (KPIs) and targets. As the initial CSRD reports are due in 2025 for the year-end 2024, organizations must swiftly gather accurate data across their structure to meet new assurance requirements. To streamline this process, GenAI automates various tasks, replacing manual processing. Generative AI is employed to search, collect, process, and analyze all necessary documents for CSRD reporting. Following the AI-driven parsing of relevant information, CSRD responses are automatically generated, calculated, visualized, and the corresponding report is published. The entire generation process is transparent to users for validation and explainability, right down to the information source.

Corporate Sustainability Reporting by NTT DATA

TCS 

EngageX​

EngageX aims to bridge the gap between HR Managers and field service workers, focusing on engagements with deskless workers. Using Generative AI, the solution deciphers and pieces together information that may be challenging to find or inaccessible, particularly concerning issues like illness, incidents, diseases, or bereavement. An Empathy Score is generated based on prompts about injury type, severity, and the number of recovery days, serving as a foundation to drive engagement empathy. This tool empowers HR managers and Project Managers to enhance employee engagement in various situations, promoting unbiased support, empathy-driven conversations, and timely interventions to foster a healthy relationship between employees and the organization.

EngageX by TCS

Coaching and Support for Partners by SAP

In the first special edition Hack2Build on Generative AI, all the partners received coaching on the new capability in SAP AI Core and early access to Generative AI Hub by SAP. The SAP Co-Innovation Lab, in collaboration with support from SAP AI Product Engineering and SAP BTP Architects from Partner Solution Adoption, ensured that the prototypes developed in Hack2Build were ready for showcase at SAP TechEd2023, both in Bangalore and virtually.

Panel of Judges

The Panel of Judges, passionate about the topic, provided valuable feedback to encourage partners in building their complementary solutions for customer adoption of SAP Business AI. Gratitude is extended to the judges for dedicating considerable time to evaluate each partner solution during the demo-jam held on September 12, 2023.

• Sebastian Wieczorek – Vice President AI Technology & Global Lead of AI Ethics
• Rudi Held – Vice President & Head of Global Co-Innovation Lab Network, PES
• Jana Richter – Vice President AI Product Success
• Pavel Penaz – Head of SAP BTP Strategy & Execution, Partner Solution Adoption

Way Forward

Looking ahead, Partners would have further interests to leverage the [recently announced](https://www.sap.com/events/teched/news-guide/ai.html#article8) SAP HANA Cloud vector for private business data access and explore the potential of the Joule generative AI copilot for extensibility. Partners have shown keen interest in these advancements, laying the foundation for the upcoming edition of [Hack2Build](https://events.sap.com/hack2build/en/calendar). SAP remains committed to supporting existing partners and extends a welcome to new ones interested in co-innovating with SAP on Generative AI in 2024. The collaborative journey continues, aiming for further advancements and partner innovations to emerge in 2024 and beyond.