? Innovating with SAP's Generative AI Hub and Azure OpenAI
Science & Technology
Introduction
Hello and welcome to the last Math Thursday of this year's Dev Toober Fest! Today, we have an action-packed session featuring our partners from Microsoft and our wonderful colleague Tanvi, who leads internal AI enablement. In this session, we’ll explore innovative AI solutions using SAP's Generative AI Hub in collaboration with Azure OpenAI Services.
Introduction
We have a comprehensive agenda today. First, Hulga Bage from Microsoft will introduce OpenAI Services and demonstrate how you can access Azure OpenAI models via the Generative AI Hub. He will provide real-world use cases, which will be followed by Tanvi’s overview of our internal AI enablement hackathon. Finally, we will hear from the three winning teams of the hackathon, who will share their experiences developing with large language models, particularly those from the GPT family.
OpenAI Services and Generative AI Hub: A Deep Dive
Hulga kicked off the session, expressing excitement about the collaborative spirit of Dev Toober Fest. He introduced himself as a Cloud Success Architect and described the integration of SAP's AI technologies with Microsoft’s offerings.
Hulga emphasized two primary components in the collaboration:
- Jewel and CoPilot Integration: Jewel, SAP's digital assistant, and Microsoft's CoPilot will together enhance user experiences by simplifying data accessibility.
- Generative AI Hub: This is SAP’s platform that connects developers to various large language models (LLMs). The Gen AI Hub enables quick access to different AI capabilities, fostering an easy integration process with Azure OpenAI Services.
Utilizing Generative AI
Hulga demonstrated how to use the Generative AI Hub effectively, showing how developers can build applications using the Azure OpenAI Services. He illustrated the AI launchpad, which allows seamless interactions with multiple models.
By working through a crossword puzzle challenge created for Dev Toober Fest, he demonstrated how different models yield different outputs. As he experimented with GPT-4 and its mini versions, he pointed out how the quality and appropriateness of responses can vary, emphasizing the importance of model selection based on tasks.
AI Learning and Hack Days
Following Hulga's session, Tanvi took the stage to discuss the Global Generative AI Learning and Hack Days initiative. This initiative was framed as an opportunity for SAP employees—regardless of their background—to enhance their understanding of generative AI. Notably, over 10,000 employees participated, submitting an impressive 700+ ideas.
The initiative aimed to transition participants from merely being aware of generative AI to actually becoming practitioners. Tanvi explained that participants formed teams, received guidance from experts, and developed prototypes using the infrastructure provided, including the Generative AI Hub.
Ultimately, three teams stood out during the hackathon. The winning projects included:
- AI-assisted RFP Responses
- Intelligent Feedback Analyzer
- Threat Modeling Automation
Each team presented their prototypes and shared insights into their experiences—motivation, challenges faced, and how they leveraged AI technologies in creative ways.
Team Presentations
AI-assisted RFP Responses: This team created a tool that automates responses to requests for proposals (RFPs) using generative AI. Their solution extracts valuable data from previous RFPs, improves response quality, and integrates it all into a cohesive format for easy export.
Intelligent Feedback Analyzer: This tool aims to improve how organizations process and analyze employee feedback, enabling better responses to initiatives and boosting engagement.
Threat Modeling Automation: Focused on enhancing security processes, this team developed a solution that simplifies threat modeling, making it accessible and less intimidating for developers.
Conclusion
Today's session highlights SAP's commitment to leveraging generative AI in innovative ways. The combination of the Generative AI Hub and Azure OpenAI presents limitless possibilities for enhancing business processes and improving efficiency. The enthusiasm and creativity showcased by our hackathon participants demonstrate the immense potential of AI in transforming how we work.
Keyword
Generative AI, Azure OpenAI, SAP, Jewel, CoPilot, AI Hub, Hackathon, Large Language Models, RFP automation, Intelligent Feedback Analyzer, Threat Modeling Automation.
FAQ
Q1: What is the Generative AI Hub?
A1: The Generative AI Hub is SAP’s platform that connects developers with various large language models, facilitating the integration of AI capabilities into applications.
Q2: How many employees participated in the Global Generative AI Learning and Hack Days?
A2: Over 10,000 SAP employees participated in this initiative.
Q3: What were the winning projects from the hackathon?
A3: The winning projects were AI-assisted RFP Responses, Intelligent Feedback Analyzer, and Threat Modeling Automation.
Q4: What technologies were used in the hackathon projects?
A4: Participants utilized Azure OpenAI services and various generative AI models during the hackathon.
Q5: Can anyone participate in initiatives like the Global Generative AI Learning and Hack Days?
A5: Yes, these initiatives are designed to be inclusive, allowing employees from all backgrounds to join and learn.