SAP AI

What is AI?

• a field of computer science that focuses on creating machines that can perform tasks that typically require human intelligence, such as learning, problem-solving, and decision-making.

SAP AI

SAP AI Strategy:-

• SAP is dedicated to creating intelligent, sustainable, and interconnected enterprises by integrating AI technologies into applications and business scenarios. This mission is achieved through the AI offerings of SAP Business Technology Platform (BTP).

How AI can help:-

• AI is essential to ensure the success of your company in the future because it improves client experiences and frees people from repetitive jobs through automation.
• Monitoring and analyzing existing processes.
• Optimizing and automating processes.
• Enabling new processes.

Why we used it?

• Cost-effective.
• Time Saving.
• Increase productivity.
• increase creativity.

How SAP is infusing AI into applications?

• SAP strategically selected artificial intelligence (AI) as a key pillar for enabling intelligent organizations. To quickly take advantage of everything that AI has to offer, SAP’s standard software, such as SAP S/4HANA, can be seamlessly integrated with AI.
• SAP provides enterprise-grade AI capabilities on the SAP Business Technology Platform (BTP) for integrating AI into business applications. This makes it possible to easily expand and improve application capabilities.

SAP AI Core:-

• It’s like the engine for your AI applications and services.
• You can teach and use AI models efficiently and in a budget-friendly way, even when you need a lot of them.
• These AI models can easily be added to your regular business apps thanks to a standard AI interface.
• You can create AI applications using any open-source tool for machine learning.
• SAP AI Core takes care of the whole AI process, from starting the project to sharing the results. It’s based on modern open-source tools like Argo Workflows and KFServing.

SAP AI Launchpad:-

• Think of it as your control center for AI.
• You can connect it to different parts of your AI system, including SAP AI Core.
• It helps you manage all your AI projects in one place.
• You can use SAP AI Launchpad to watch how your AI models are doing and make them even better over time.

Phases of an AI Process:-

SAP AI Core:-

What is the added value of SAP AI Core?

• Peace of Mind: SAP AI Core takes care of everything you need for AI, so you don’t have to worry about it.
• Easy to Use: It’s simple to use because it hides all the complicated stuff and gives you easy ways to use AI in your business apps.
• Fast Integration: You can quickly add AI to your SAP apps and business processes.
• Cost Control: It helps you balance your AI costs and performance. You can speed up your AI when you need it and slow it down when you don’t, which can save you money.
• Flexible Infrastructure: SAP AI Core uses something called Kubernetes, which is fast and flexible. It can adjust the resources for your AI as needed. For example, it can give more power to heavy AI tasks or complex AI jobs.

Model Reuse:-

• Simplified Training: SAP AI Core makes it easy to create and use AI models. You can even share model templates with others so they can train the models with their data.

What SAP AI Launchpad Does:-

• Central Management: It’s like a control center for AI on SAP BTP, helping you keep track of all your AI models.
• Easy Access: You can connect it to different AI systems, including SAP AI Core.
• Streamlined Management: It makes managing AI projects easier with a user-friendly interface.
• Metrics and Analysis: You can see and analyze how well your AI is performing, and it’s good for keeping an eye on important data.
• Standardized Connection: It connects with AI systems through something called “AI API,” making everything work smoothly.
• Full Lifecycle Support: It helps you with the whole AI process, from start to finish, and lets you check important AI measures.

SAP AI Launchpad:-

• SAP AI Launchpad is a service on SAP Business Technology Platform.
• It manages AI models transparently across the enterprise.
• It connects to AI API-enabled runtimes like SAP AI Core.
• It centralizes AI lifecycle management with a user-friendly interface.
• It acts as a single access point for all AI content in SAP.
• It captures and analyzes metrics from supported AI runtimes, such as SAP AI Core.
• Users can compare and visualize these metrics.
• Integration with AI runtimes is done through a standardized interface called “AI API.“
• It supports the full lifecycle management and operations of AI processes.
• Customers can productize existing training models or trigger jobs and deploy models using SAP AI Launchpad.

Model Training in SAP AI Core

Model Deployment in SAP AI Core:-

• Model Deployment in SAP AI Core is essential after model training for predictions.
• It’s known as Model Serving.
• Kubernetes clusters support both CPU and GPU containers for Model Serving.
• Kubernetes allows scaling the Model Server to handle many requests efficiently.
• Two scaling options: Autoscaling adds containers as needed, and Scale to Zero saves costs by shutting down idle containers.
• Deploying a model involves creating a web app that serves predictions through an internet-accessible endpoint, easily scalable on Kubernetes infrastructure.

1. Autoscaling:-
   •Description: Autoscaling is a dynamic process where new containers are automatically created (cloned) or terminated based on the current workload or demand. It’s typically used to ensure that your application can handle increased traffic or processing requirements without manual intervention.
   •Example: Imagine you have a web application that provides stock market data. During regular hours, the traffic is moderate, and your application runs on a few containers. However, when the stock market opens, there’s a sudden surge in users accessing your app to get real-time data. Autoscaling detects this increased demand and automatically adds more containers to handle the extra traffic. When the stock market closes and demand decreases, it scales down by removing unnecessary containers to save resources.
2. Scale to Zero:-
   •Description: Scaling to Zero is a cost-saving strategy that involves shutting down idle or unused containers when they are no longer needed. This approach helps optimize resource utilization and reduce expenses, particularly in cloud computing environments where you pay for the resources you use.
   •Example: Consider a company that offers an online training platform. During the day, many users access the platform, and containers run to handle the load. However, at night, the platform experiences significantly lower user activity. With Scale to Zero, the system detects the reduced demand and automatically shuts down containers that are no longer in use. This means the company only pays for the resources it actually needs when there is active usage, saving costs during periods of low activity.

In summary, Autoscaling is about performance and responsiveness, while Scale to Zero is about cost optimization. Both are valuable techniques, but they serve different purposes in managing containerized applications.

I hope it’s useful.

Thank you, 😊.