Menu

niranjan

About niranjan

Posts by :

Automating Product Classification in ECommerce with SAP AI Services

Automating Product Classification in ECommerce with SAP AI Services

Introduction

The SAP AI Data Attribute Recommendation service automatically analyses your data and helps to predict the missing information or organizing it based on patterns it has learned from similar data.

For example, imagine you have a lot of product information coming in— Data Attribute Recommendation service can automatically help categorize these products and even predict the prices based on their descriptions.

It is leveraging machine learning to analyse existing data patterns and generate attribute suggestions. It is reducing the manual effort required for data classification and improving data accuracy and consistency across business processes.

Note: SAP AI Data Attribute Recommendation service is part of SAP AI Services under   SAP AI Foundation. It is covered in our course about “SAP AI Services”.

What You Will Learn 

In this tutorial, we will walk through a real-world use case to demonstrate the power of AI in business applications: Automating Product Classification with AI Models. 

Scenario: 

In the world of ecommerce, assigning the correct product categories and hierarchies when creating new product data is typically a manual, time-consuming task. This task often requires expert knowledge to ensure accuracy, and as businesses scale, the challenge only grows. 

Challenges: 

Manual product classification can lead to delays, errors, and inefficiencies. As the product catalog expands, the need for an automated solution becomes essential to ensure speed and consistency. 

Solution: 

In this tutorial, we’ll demonstrate how AI models, specifically the Data Attribute Recommendation Service, can streamline this process.  

By leveraging AI to analyze historical data, the system automatically suggests the most appropriate product categories and attributes with a confidence score. This not only reduces manual effort but also improves accuracy and efficiency. 

This approach can also be applied to other use cases, such as predicting attributes for sales orders or filling in missing data in transactions. 

What You’ll Do: 

We’ll guide you step by step through an end-to-end tutorial where you’ll: 

  • Upload a product dataset schema. 
  • Train an AI model using historical data for product classification. 
  • Generate predictions with confidence scores for new products, automating the process of assigning categories. 

Lets dive in! 

 Prerequisites 

You need an SAP BTP Free Trial Account. No need of a paid SAP BTP Account. 

Setting Up of SAP Data Attribute Recommendation Service 

Step 1: Logon to your SAP BTP free trial account. 

Step 2: Click on Boosters from the Gobal Account view. 

Step 3: Search for Data Attribute Recommendation and click on it. 

Step 4: Press Start. 

Step 5: Download these Postman files. We will be using it in the future steps. 

Install Postman REST Client 

Step 1: Visit official Postman website and download postman application. Postman client is required to test the Data Attribute Recommendation Service through APIs. 

This is how Postman UI looks like once you install it on your computer and open it. 

Import Postman Environment and API Collection  

Step 1: Click on Import button.  

Step 2: Import both Postman Environment and Postman Collection files. 

Step 3: Press on Import. 

Step 4: Next, tap on Settings icon. 

Step 5: Disable SSL certificate verification. 

Step 6: Enable ”Automatically follow redirects”. 

Step 7: Lastly, choose Environment as “Data Attribute Recommendation Tutorial”. 

Obtain Access Token 

Step 1: On the left pane Expand navigate to Data-Attribute-Recommendation Demo > Setup > 000.Get Authorization. Click on Send. 

If the request is successful. A token will Generated as shown below. 

Creating a Dataset Schema 

Step 1: Navigate to Data-Attribute-Recommendation Demo > Upload Data > 002.create new Dataset Schema. Press Send. 

If the request is successful it will show “200 OK”. A schema is successfully created.

Step 2: Then navigate to 001.Get Dataset Schema. Click Send. It will show the schema we created in the previous step.  

Here Details can be accessed at anytime.

Creating a Dataset 

Navigate to 005.Create new Dataset. Click on Send. A dataset is a table where we will upload all our data. 

If it was successful, it will show a “200 OK” message.

Download sample data available on GitHub. Download the CSV file that contains the data.

Upload Sample Data for Machine Learning Training 

Step 1: Navigate to 006.Upload Data 

Step 2: Choose Binary. And click on “New file from local machine”.

Step 3: Upload the sample excel file provided with the course to train the Machine Learning model. It contains 1000s of records. Based on the patterns ML model will get trained.

Step 4: Click on Send. 

If it throws the message “200 OK” validation has started. You will be able to see that status of data validation under property “status”. 

Step 5: To check status of the data validation of the file uploaded navigate to 007.Get Dataset by id and click on Send. The status must change to Succeeded as given below. 

Model Training using Sample Data 

Step 1: Navigate to Train job > 002.Create new Job. Click on Send. 

Initially the status will be “Pending” and in a few minutes it will change to “Succeeded”. 

Step 2: To check the status of the training job, navigate to Train Job > 003.Get Job by id. Click on Send. 

If status show Succeeded, the process has complete. 

Deploy the Machine Learning Model 

Step 1: Navigate to Deploy Model > 002.Deploy Model by name. Click on Send. 

Step 2: To check status of deployment navigate to Deploy Model > 003.Get Deployment by id. Click on Send. Wait for a few minutes to change status to “Succeeded”.

Testing the Machine Learning Model 

Step 1: Navigate Classify Records > 001.Send inference request.  

In the body of the request a sample data is given. Based on the model the data needs to be categorized by the ML model that we’ve just trained. 

 Step 2: Click on Send.

If the response is “200 OK” the request was successful.  

In the response body of the request, you will be able to see the predictions made by the ML model that we’ve trained with probability.

Conclusion 

Machine learning has revolutionized the way organizations classify and manage data, offering a scalable and efficient approach to handle large datasets. By following the steps outlined in this blog—preparing your data, choosing the right ML algorithm, training your model, and evaluating its performance—you can create a reliable data classification system tailored to your specific needs.  

With SAP AI, you can take your data classification efforts to the next level by integrating cutting-edge AI capabilities into your existing systems.  

SAP Integration Suite Meets SAP AI: Accelerating iFlow Creation with Generative AI

SAP Integration Suite Meets SAP AI: Accelerating iFlow Creation with Generative AI

,

SAP Integration Suite is a cloud-based integration platform that enables businesses to seamlessly connect and integrate applications, data, and processes across diverse systems, both on-premise and in the cloud. 

It has introduced a groundbreaking new feature: Generative AI Based Integration Flow Generation, powered by SAP AI.

This functionality harnesses the power of SAP Artificial Intelligence to streamline and accelerate the creation of integration flows, allowing users to simplify complex workflows with ease.

By automating the process of building these iflows with SAP AI we can reduce the need for manual intervention, automate tedious tasks, reduce errors, and increase productivity of consultants.

This marks a significant leap forward in how SAP Integration Suite supports AI-driven innovations for seamless and scalable integrations. 

What is AI-Assisted Generation of Integrations?

SAP Integration Suite is offering a powerful AI feature in the suite to create  integration flows(iflows). Once the feature is activated, SAP integration consultants can leverage AI to generate integration flows.

The consultant simply provides a description of the integration scenario when prompted, and the Generative AI tool uses that input to automatically create the corresponding integration flow.

The AI-driven approach is simplifying the development process, enabling faster and more efficient creation of integrations based on specific requirements.

Let’s dive in and create our first integration flow using Artificial Intelligence.

Enable Generative AI Feature

Step 1: Logon to the integration tenant.

Step 2: Go to Settings>Integrations>Generative AI. And choose Edit. 

Enable the Activate Generative AI Features checkbox.

Next, click Save.

Generate Integration Flows with AI

Step 1: You need to add relevant systems like S/4Hana under System Landscape in SAP BTP Cockpit.

Once a successful SAP Integration Suite formation is established with relevant systems, the Generative AI-Based Integration Flow Generation feature automatically scans the systems and their APIs.

The systems and APIs will be visible in the Sender and Receiver sections of the Generative AI-based integration flow creation UI once the above is configured. This capability makes it much easier to select and link systems, simplifying the entire process of building iflows.

For detailed guidance on enabling and configuring System Landscape in SAP BTP Cockpit for SAP Integration Suite, please refer to the documentation. 

Step 2: From the Integration tenant Navigate to Integrations and APIs > Edit.

Step 3: Click on Add > Integration Flow.

Step 4: Choose Generate Integrations with assistance from AI.

Step 5: Describe the integration scenario and Click Send. You can also choose a sender and receiver system.

Step 6: Click Generate .

Congrats. you’ve now generated an iflow using the Generative AI-Based Feature in SAP Integration Suite.

Summary

In this blog, we’ve explored the powerful new Generative AI-based integration flow generation feature in SAP Integration Suite, which leverages SAP AI to automate and streamline the creation of integration flows.

By reducing manual effort and simplifying complex workflows, this feature enhances productivity for consultants and ensures more accurate, efficient integrations.

The step-by-step tutorial provided demonstrates how easy it is to activate and use this functionality, showcasing how AI-driven innovations are shaping the future of SAP Integration Suite for seamless, scalable business operations.

Thank you, and we will see you in the next blog! 

How to Build Your First Full-Stack App in SAP Build Code with Joule and Generative AI

How to Build Your First Full-Stack App in SAP Build Code with Joule and Generative AI

, , , , , ,

In this blog, we’ll guide you through the process of building a full-stack application from scratch using SAP Build Code and SAP Joule’s Generative AI.

We’ll leverage SAP Joule’s Generative AI to streamline the development of data models and application logic for essential functionalities like customers, purchases, and redemptions. Whether you’re a seasoned developer or just starting out, this walkthrough will show you how these powerful tools can accelerate your development process and enhance your projects. Let’s dive in!

Before diving in, here’s what we’ll cover in this guide:

  • Create a full-stack project using a template.
  • Develop data entities and services with Joule’s Generative AI.
  • Generate application logic code using Joule’s Generative AI capabilities.
  • Add a UI to the application and perform testing.

Step 1: Logon to SAP BTP Trial. If you don’t an account and want to setup one this blog will help.

Once your trial account is ready follow these steps: Click on “Go To Your Trial Account”.

Step 2: Click on  Boosters.

Step 3: Search for “Get started with SAP Build Code ” and Click on it.

Step 4: Click on Start.

And, wait for few minutes to setup the SAP Build automatically for you.

Step 5: Then, click on Navigate to Subaccount.

Step 6: Once you’re inside the subaccount click on Instances and Subscriptions.

Step 7: Click on SAP Build code.

Step 8: Then, Click on Create. 

Step 9 :Choose Build an Application. Then, choose SAP Build Code.

Step 10 :Choose SAP Build Code.

Step 11: Next, choose Full Stack Application.

Step 12: Provide a name for the project of your choice and click Create. And select Node or Java as your development stack. 

Step 13:  It will create a project. Choose the project that you’ve created.

Step 14 :Click on Ok. It will navigate you to SAP Business Application Studio.

Step 15: Click on Joule Icon.

Step 16: Type “/” in the command section and select “cap-gen-app.“If Joule is throwing an error like “Joule is not enabled for you” follow these steps in SAP Community.

Step 17: Then, paste the below prompt and press send.

“Design a customer loyalty program application. Define 4 data entities: Customers, Products, Purchases and Redemptions. Each customer must have the following fields: name, email, 7-digit customer number, total purchase value, total reward points, total redeemed reward points. All fields for each customer should be integer except name and email that will be stored as string. Each product should have a name, description and price. Purchases should include the following fields: purchase value, reward points. All fields in Purchases must be integer. Redemptions must have 1 field in integer: redeemed amount. Each purchase and redemption will be associated to a customer. Each purchase will be associated to a product and is called selectedProduct. “

Step 18: Click on Accept .

Step 19: The screen will get updated once the prompt is executed. Then choose Open Editor. And select Sample Data.

Step 20:Choose Customers and Value 5 and press on Add. It will create 5 sample records for customer data.

Step 21: Click on Enhance.

Step 22: Next, click on StoryBoard.

Step 23: And select Purchases > Add logic.

Step 24: Click on Add.

Step 25: Then, choose on Open Code Editor > Application Logic.

Step 26: Copy the below prompt and paste it into the text field as given below press the send button. It will send the prompt to Joule to process.

“Reward points of each purchase will be the one tenth of the purchase value. Each purchase value will be added to the total purchase value of the related customer. Each reward point will be added to the total reward points of the related customer.”

Step 27: Click on Accept.

Step 28: Then, choose StoryBoard.

Step 29: And, select Create a UI application.

Step 30: Fill in the text fields as follows:

Display name: Purchases

Description: Manage Purchases

Then, click Next.

Step 31: Choose Templete-Based. And, click Next.

Step 32: Choose List report page. Click Next.

Step 33: Choose main entity as Purchases. Click on Finish.

Repeat steps 29-33 to create additional UI apps for the Customers and the Redemptions.

Details to be filled for Customer and Redemptions is given below:

Customer:

  • Display name: Customers
  • Description: Manage Customers
  • UI Application type: Template-Based Responsive Application
  • UI Application Template: List Report Page
  • Main Entity: Customers

Redemptions:

  • Display name: Redemptions
  • Description: Manage Redemptions
  • UI Application type: Template-Based Responsive Application
  • UI Application Template: List Report Page
  • Main Entity: Redemptions

Step 35: Fantastic! We’re ready to test the app. To test the application click on Run and Debug button.

A new tab opens. Click on Customers > Go.

It will open the list of customer data. Choose any one of the customers to test the edit function.

Click on the “Edit” and change Name and Email as given below, and Save.

Great, now you are able to see the list of customers and you can edit it too.

Let’s test the Create functionality of the app. Click on the Create button.

And fill the details for a new customer as given below.

The new customer is now added successfully.

Summary

In summary, this guide has detailed how to build a full-stack application from a template using SAP Build Code, incorporating SAP Joule’s Generative AI capabilities. We utilized Joule to generate code for data entities and services related to customers, purchases, and redemptions, and also to develop the application logic. By leveraging these tools, you can streamline the development process and create robust applications efficiently. Continue exploring SAP Build Code and Joule to unlock further possibilities and enhance your development projects.

 

Here is a detailed demo to help you learn further!

Kickstart Your Journey with SAP Artificial Intelligence: Part 4 Fundamentals of Generative AI

Kickstart Your Journey with SAP Artificial Intelligence: Part 4 Fundamentals of Generative AI

Introduction

Now that we’ve covered Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning, it’s time to explore an even more fascinating area—Generative AI. 

This 7 part blog series is designed for anyone who wants to start learning about SAP AI. No matter what your job is or what you do, understanding the basics of Generative AI how it connects to SAP is a smart choice.

In this blog series, we’ll break down SAP AI into easy-to-understand parts, covering one step at a time.

SAP AI step 4 of 7

Part 1 — Introduction to AI 

Part 2 — What is Machine Learning?

Part 3 — Basics of Deep Learning

Part 4 — Getting Started with Generative AI [Current Blog]

Part 5 — What Are Large Language Models (LLMs)?

Part 6 — Prompt Engineering: How to Communicate with AI

Part 7 — Introduction to SAP AI

What is Generative AI?

Generative AI is a form of artificial intelligence designed to create new content, such as artwork, music, or realistic images, without explicit instructions on what to produce.

Unlike traditional AI, which is task-oriented and problem-solving, Generative AI stands out for its ability to mimic human-like creativity.

It can generate original and unique content, ideas, or solutions, much like humans do.

Let’s Break it Down with a New Example

Imagine I asked you to design a vehicle that doesn’t exist. You’d need to use your imagination to invent something entirely new, something no one has ever seen before.

Thanks to our natural creativity, you might come up with a vehicle that has the body of a spaceship, the wheels of a monster truck, and the wings of a bird. It’s completely original, something you dreamed up from scratch.

Now, imagine a computer program doing something similar. Instead of just following a set of instructions, it can generate new things—like artwork, music, or even realistic images—without needing exact directions. It’s as if the program has its own creative spark, just like you.

Imagine a computer program has been trained by looking at countless aerial photos of eagles flying over mountains. With this knowledge, it can now generate a completely new image—not by copying any of the pictures it has seen, but by combining its understanding of what makes an eagle and a mountain unique. The result could be an entirely new perspective, like an eagle soaring high above rugged peaks, blending elements from what it learned to create something fresh and original.

Eagle Flying Over Mountains Generate by AI
Eagle Flying Over Mountains Generate by AI

This is Generative AI — a machine that can imagine and create on its own. It can draw pictures, tell stories, or even invent new games, all without needing anyone to show it how. It’s like giving a computer the ability to dream up new ideas, just like we do.

Where Does Generative AI Fit into the AI Hierarchy?

Generative AI is a branch of Deep Learning. It sits within a larger framework of artificial intelligence. Imagine AI as the broadest category, with Machine Learning being a more specific part of it. Within Machine Learning, we have Deep Learning, and finally, Generative AI is a subset of Deep Learning. The diagram below helps visualize how AI, Machine Learning, Deep Learning, and Generative AI are related.

Artificial-Intelligence-Hierarchy

Generative AI builds on machine learning techniques, particularly deep learning and neural networks.

What makes Generative AI stand out is its ability to create brand-new content.

Traditional AI, machine learning, and even deep learning have largely focused on predictive models—used to recognize, classify, or forecast patterns in data. For example, a typical machine learning task might be identifying a car in a set of images or grouping different vehicles based on their features.

Generative AI goes further. Instead of just identifying a car, it can create an entirely new design of a futuristic vehicle or imagine and generate something that’s never been seen before. This creative capability was once thought to be unique to humans.

The image below showcases AI’s evolution over time. The leap from rule-based systems to Generative AI has been powered by advances in learning algorithms, greater computational power, and access to vast amounts of data.

Evolution of AI with Time

What are Generative Models?

Generative Models are the heart of Generative AI. These machine learning models learn the patterns and structures of existing data, then use that knowledge to generate new, original content.

Applications:

  • Visual Arts: Creating stunning images and art pieces.
  • Music Composition: Generating unique musical compositions.
  • Text Generation: Writing articles, stories, and scripts.

Common Generative Models:

  • Variational Autoencoders (VAEs): Known for their versatility.
  • Generative Adversarial Networks (GANs): Creating highly realistic content.
  • Restricted Boltzmann Machines (RBMs): Often used for collaborative filtering.
  • Transformer-based Language Models: Excelling in natural language tasks.

We’ll explore these models in detail and see how they’re revolutionizing various industries.

Usages of Generative AI in Real-life

Here are some examples of how Generative AI is being used to create real-life applications:

Generative AI Use Cases

Software:

  • Predictive Maintenance: Using AI to predict when equipment will fail, reducing downtime and costs.
  • Virtual Assistants: AI-powered assistants that can handle customer inquiries and tasks, improving efficiency.

Content Creation:

  • Creative Writing: AI can generate creative text, such as poems, scripts, or marketing copy.
  • Art Generation: AI can create original artwork in various styles, from abstract to realistic.

NLP(Natural Language Processing):

  • Language Translation: AI can translate text and speech between different languages, facilitating global communication.
  • Sentiment Analysis: AI can analyze text to determine the sentiment expressed (e.g., positive, negative, neutral), providing valuable insights for businesses.

Here are Some Popular Generative AI Tools

There are plenty of tools available today in the market, here are the popular ones you can try.

  1. ChatGPT: Developed by OpenAI, ChatGPT engages in natural language conversations, generating contextually relevant and coherent responses using deep learning techniques.

  2. AlphaCode: Created by DeepMind, AlphaCode generates optimized code solutions across various programming languages, leveraging deep learning to analyze and learn from code patterns.

  3. GitHub Copilot: An AI-powered code completion tool by GitHub and OpenAI, integrated into code editors to provide real-time suggestions and completions.

  4. Gemini: Google’s conversational AI chatbot, initially based on LaMDA and later upgraded to models like PaLM and Gemini.

  5. Microsoft Copilot: An AI-powered assistant by Microsoft for tasks like summarizing articles, reformatting text, and updating images.

  6. DALL-E: Developed by OpenAI, DALL-E generates images from text descriptions, with versions including DALL-E2 and DALL-E3.

  7. StyleGAN: Created by NVIDIA, StyleGAN generates high-quality synthetic images, especially realistic human faces, with control over visual features.

An image generated using StyleGAN that looks like a portrait of a young woman. Generated by an artificial neural network based on an analysis of a large number of photographs.
An image generated using StyleGAN that looks like a portrait of a young woman.

Summary

In this blog, we explored what generative AI is and how it differs from other types of AI. We also discussed several real-life applications of generative AI and highlighted some popular tools in this field.

Kickstart Your Journey with SAP Artificial Intelligence: Part 3 -Basics of Deep Learning

Kickstart Your Journey with SAP Artificial Intelligence: Part 3 -Basics of Deep Learning

Introduction

By now, you’ve already learned about Artificial Intelligence (AI) and Machine Learning (ML). You understand how AI can help a robot find your lost toy and how ML lets it learn from examples. Now, we’re going to dive even deeper into a special kind of Machine Learning that is Deep Learning. 

So, if Machine Learning is like teaching a robot to recognize your toys, then Deep Learning is like giving that robot a superpower—it can learn to recognize anything by itself, just by looking at lots of examples, even without you having to explain what each thing is. Let’s explore how this works! 

This 7 part blog series is designed for anyone who wants to start learning about SAP AI. No matter what your job is or what you do, understanding the basics of Generative AI how it connects to SAP is a smart choice.

In this blog series, we’ll break down SAP AI into easy-to-understand parts, covering one step at a time.

SAP AI step 3 of 7

Part 1 — Introduction to AI 

Part 2 — What is Machine Learning?

Part 3 — Basics of Deep Learning [Current Blog]

Part 4 — Getting Started with Generative AI

Part 5 — What Are Large Language Models (LLMs)?

Part 6 — Prompt Engineering: How to Communicate with AI

Part 7 — Introduction to SAP AI

What is Deep Learning?

Imagine if a machine could learn just like you do! That was the big idea behind creating something called “Deep Learning.”

Deep Learning is a special kind of machine learning, which is part of something even bigger called Artificial Intelligence (AI). It’s like teaching a computer how to think a little bit like a human brain!

SAP AI -Deep Learning

Deep learning uses something called an Artificial Neural Network (ANN), which works a bit like how our brains connect and think. These networks help machines recognize patterns, solve problems, and even make decisions, just like we do when we learn new things!

First, lets understand a few important concepts,

Biological Neural Network in the Human Brain

The human brain is like a supercomputer! Its most important building blocks are called neurons. There are billions of neurons in our brain, all connected, forming a neural network. Think of it like a giant web of wires, where each wire passes signals between different points.

Neurons take in inputs from everything we experience—what we see, hear, or even feel—and send messages to other neurons. These messages help us think, react, and most importantly, learn new things!

Artificial Neural Network (ANN)

Let’s simplify an Artificial Neural Network by imagining a classroom of students working together on a big project. Each student in the class represents a neuron, and they all have to talk to each other to complete the project.

Each student gets some information (like a task or a question) from the teacher. They work on it and then pass their answer to the next student. Some students do simple tasks, while others work on more complex parts. In the end, the final student collects all the information and presents the complete solution.

The information the students pass around is like data, and the students are like neurons in the ANN.

Architecture of an Artificial Neural Network

Now, let’s break down the layers of this “classroom network”:

  1. Input Layer: This is like the teacher handing out a task to the students. It’s where the network gets its initial information, such as an image, sound, or data.
  2. Hidden Layers: These are the students in the middle who discuss and solve parts of the task. They don’t talk directly to the teacher but share their work with each other. These layers are where most of the problem-solving happens, just like how the brain thinks through a problem.
  3. Output Layer: This is like the final student presenting the finished project to the teacher. It gives the answer, whether it’s identifying a picture or making a prediction.

In a Nutshell:

  • Input Layer: Where the network receives the problem or data to solve.
  • Hidden Layers: Where the network thinks, processes the data, and finds patterns, like students discussing and solving tasks together.
  • Output Layer: Where the network gives its final answer or decision, like a student presenting the solution.

By thinking of an ANN as a classroom where students (neurons) collaborate to solve a problem, it becomes easier to understand how deep learning works!

How Does an Artificial Neural Network Work?

Imagine a group of students trying to identify a type of fruit by sharing their observations.

Each student focuses on specific features, such as color, shape, and texture.
Individually, they might not fully figure out what fruit it is, but by combining their insights, they can arrive at a collective understanding.

In the world of artificial neural networks, these students represent neurons.

In an artificial neural network, individual neurons (like the students) specialize in recognizing specific aspects of the input. When combined, they contribute to recognizing the overall object (like the fruit). The network improves over time, just as the students get better at identifying fruits with more practice.

Input Layer (Observation):

Each student observes one detail—such as the color, roundness, or smoothness of the fruit—forming the input layer of our network.

Hidden Layers (Processing):

The students then discuss their observations with each other, simulating the hidden layers of a neural network. By sharing and refining the information, they build a more complete picture of the fruit’s features.

Output Layer (Recognition):

Finally, they reach a conclusion by combining all the details. If most agree that the fruit is an apple, they say, “It’s an apple!” This corresponds to the output layer, where the network gives its final decision.

Scoring Approach:

To improve their fruit recognition skills, the students keep track of their accuracy.

If they correctly identify an apple, they earn points. If not, they learn from their mistakes and adjust their approach next time. Similarly, in a neural network, a scoring approach is used to adjust the network’s parameters to improve its accuracy over time.

This teamwork illustrates how artificial neural networks process information step by step, learning from various features and refining their understanding through each attempt.

Deep Neural Networks

A Deep Neural Network (DNN) is a type of Artificial Neural Network (ANN) that has multiple layers between the input and output.

The term “deep” means it contains several layers, which allows it to learn more complex patterns compared to a simpler network.

Key Points about Deep Learning

Let’s summarize some important concepts about Deep Learning:

Subset of Machine Learning (ML)

Deep learning is a part of machine learning, which itself is a subset of artificial intelligence (AI). Think of AI as the largest circle, with machine learning and deep learning as smaller circles within it.

Inspired by the Human Brain

Deep learning uses artificial neural networks, which are designed based on how the human brain works. Just as our brains process information through neurons, these networks process data through nodes.

Artificial Neural Networks (ANN)

An ANN is a computational system that mimics the way biological neural networks function in the human brain. The network consists of layers of interconnected nodes (like neurons) that pass information to each other.

Deep Neural Networks

The term “deep” refers to the multiple layers used in the network. Deep neural networks use several hidden layers, which allow the system to understand more complicated patterns, like recognizing an object in a photo or translating languages.

Learning from Data

Deep neural networks learn by being exposed to large amounts of data. The network adjusts the connections between its nodes based on how close its predictions are to the correct answers. For example, if the network is trying to identify animals in pictures, it learns to improve with each image it sees.

Solving Complex Problems

Deep learning is especially useful for handling difficult tasks, such as recognizing speech, analyzing images, or detecting anomalies, where traditional methods may struggle to succeed.

Machine Learning vs Deep Learning

  • Machine Learning (ML) is a subset of AI that focuses on enabling machines to learn from data without being explicitly programmed. It typically relies on simpler algorithms and may require manual feature extraction.
  • Deep Learning (DL), on the other hand, is a more advanced subset of ML, utilizing multi-layered artificial neural networks to automatically extract and learn features from large datasets, making it highly effective for complex tasks like image recognition and natural language processing.

SAP AI comparison between Machine Learning and Deep Learning

Summary

In this blog, we explored what deep learning is and how it functions. Deep learning gets its name from the multiple layers in its neural networks, similar to the way the human brain processes information through layers of interconnected neurons. Each layer adds to a deeper understanding of the data it handles.

From recognizing images and understanding speech to powering voice assistants and self-driving cars, deep learning has proven to be highly effective at solving complex problems across various fields.

Kickstart Your Journey with SAP Artificial Intelligence: Part 1 – Introduction to AI

Kickstart Your Journey with SAP Artificial Intelligence: Part 1 – Introduction to AI

, , , , ,

Introduction

The term “Artificial Intelligence (AI)” was first used back in 1956, but it was mostly discussed in scientific research or shown in movies. However, with the rise of ChatGPT, AI has become a hot topic everywhere.

This 7 part blog series is designed for anyone who wants to start learning about SAP AI. No matter what your job is or what you do, understanding the basics of Generative AI how it connects to SAP is a smart choice.

In this blog series, we’ll break down SAP AI into easy-to-understand parts, covering one step at a time.

SAP AI Srep 1 of 7

Part 1 — Introduction to AI [Current Blog]

Part 2 — What is Machine Learning?

Part 3 — Basics of Deep Learning

Part 4 — Getting Started with Generative AI

Part 5 — What Are Large Language Models (LLMs)?

Part 6 — Prompt Engineering: How to Communicate with AI

Part 7 — Introduction to SAP AI

 

Artificial Intelligence (AI) — A Kid’s Perspective

Let’s break down AI in a simple way. Imagine you’re looking for your favorite lost toy.

Here’s what you need to do to find it:

1. Identify Your Toy: You need to know exactly what your toy looks like.

2. Recognize Similar Items: If you find other toys, you need to tell if they’re the same as your lost toy. If it is the same type, you’ll check if it’s your specific toy.

3. Plan Your Search: You should come up with a plan to find your toy. For example:

  • First, check your room.
  • If it’s not there, look in the living room where you play with your toys.
  • If you still don’t find it, ask your family if they’ve seen it.

4. Adapt to the Situation: If your room is messy, you might need to search carefully among the clutter.

Now, imagine someone tells you, “I saw your toy in the backyard.”

Here’s what you would do:

  • You know where the backyard is and how to get there.
  • You won’t mistake a different toy for your own.
  • When you see the toy, you’ll check if it’s your favorite one.

You can find your toy because you have all these smart abilities.

But what if we could give these abilities to a robot so that the next time you lose your toy, the robot could help you find it?

Imagine the robot can move around and take pictures. That’s a good start, but it needs to think and act like you do. For example:

1. Identify the Room: The robot should recognize your room even if the furniture is rearranged.
2. Spot Your Toy: The robot should be able to identify your specific toy from other toys.
3. Understand Instructions: The robot needs to understand what you say and follow your directions.
4. Plan and Adapt: The robot should be able to make a search plan and adjust based on the situation, like searching more thoroughly if the room is messy.

In short, to find your toy, the robot needs to have human-like intelligence.

So Artificial Intelligence (AI) is — creating smart, human-like abilities in robots, machines, or computers.

What is AI?

Artificial Intelligence (AI) is when machines or computers are designed to imitate how humans think and make decisions.

In simple terms, AI is about making computers capable of “thinking” like humans.

AI allows computers to understand and analyze data, and make decisions on their own without needing constant human help. These smart machines use algorithms—detailed sets of instructions—to process information and get better at tasks over time.

Everyday Uses of AI

You might be using AI every day without even realizing it! Whether it’s the smart features in your apps or the technology behind your favorite services, AI is all around us, making life more convenient.

Here are some common ways AI is used in our daily lives:

Personalized Fitness Apps

Fitness apps like MyFitnessPal or Fitbit use AI to track your exercise routines and dietary habits. They analyze your data to offer customized workout plans and nutritional advice tailored to your personal goals and progress.

Navigation and Traffic Apps

When you use apps like Google Maps or Waze, AI helps find the fastest route to your destination. These apps analyze real-time traffic data to provide updates on road conditions and suggest alternative routes to avoid delays.

Smart Home Devices

Smart thermostats like Nest and security cameras like Ring use AI to learn your preferences and adjust settings automatically. For example, a smart thermostat can learn when you’re home or away and adjust the temperature to save energy while keeping you comfortable.

Email Filtering and Organization

Email services such as Gmail use AI to filter out spam and categorize your messages. AI helps organize your inbox by sorting emails into different categories, like primary, social, and promotions, so you can easily find important messages.

Streaming Music Services

Music streaming platforms like Spotify use AI to create personalized playlists based on your listening habits. AI analyzes the songs you play and suggests new music you might enjoy, enhancing your listening experience.

Voice-to-Text Apps

When you use speech-to-text features on your phone or computer, AI converts your spoken words into written text. This technology is useful for composing messages, taking notes, or creating documents hands-free.

Online Travel Booking

Travel websites like Expedia or Booking.com use AI to recommend flights and accommodations based on your past searches and preferences. AI helps you find deals and options that match your travel needs and budget.

These examples illustrate how AI is integrated into our daily routines, making tasks easier and more efficient while often working quietly behind the scenes.

Understanding AI Types

Let’s break down the types of AI in a way that’s easy to understand. Imagine AI as different kinds of robots or programs that can do different things. Here are the main types:

1. Narrow AI

Think of Narrow AI as a robot that is really good at doing just one specific job. For example, imagine a robot that can only play chess. It’s amazing at chess but can’t do anything else, like make sandwiches or tell jokes. Most of the AI we use today is Narrow AI. For instance, Siri on your phone can help with setting reminders or telling you the weather, but it can’t do your math homework or walk your dog.

Example: A calculator is a type of Narrow AI. It’s great at doing math, but it can’t help you with writing a story or making a drawing.

2. General AI

General AI is like having a super-smart robot that can do almost anything a human can do. It can play chess, help with homework, cook dinner, and even understand jokes. Right now, we don’t have General AI. It’s more like something you might see in movies where robots are almost as smart and versatile as humans.

Example: Imagine a robot that can help you with school projects, play games, and even chat with you about your favorite TV shows. This robot would be able to learn and do many different things, just like a person.

3. Artificial Superintelligence (ASI)

This is the most advanced type of AI, which is smarter than any human in every possible way. ASI would be able to solve complex problems, invent new things, and make decisions better than anyone else. But, this kind of AI only exists in science fiction for now.

Example: Think of a robot that could solve any puzzle in the world, create incredible art, and invent new technologies that we haven’t even thought of yet. It’s like having a super-genius robot who knows everything!

Generative AI fits under Narrow AI

Generative AI is a special kind of Narrow AI that can create new things. Instead of just following rules or doing tasks, it can make new content, like writing stories, drawing pictures, or composing music. For example, a Generative AI might help you write a fun story about a dragon and a spaceship, or it could generate a new piece of music based on your favorite songs.

Example: Think of an AI that helps you design a new video game level or comes up with new ideas for a storybook. It uses what it knows to create something new and original.

Generative AI fits under Narrow AI because it specializes in creating new content and isn’t yet capable of the broad range of abilities that General AI would have.

AI vs. Human Intelligence: Key Differences

Artificial Intelligence (AI) and human intelligence each have their unique strengths and characteristics. While AI is remarkable at handling numbers and rules with remarkable speed and accuracy, human intelligence brings emotional depth, creativity, and adaptability. Think of AI as a super-efficient calculator, whereas human intelligence is like a vibrant, ever-evolving work of art!

Gen AI Zequance AI

Here’s a closer look at how AI and human intelligence differ:

How They Learn:

  • AI: Learns by analyzing vast amounts of data and identifying patterns. It becomes proficient in specific tasks by processing numerous examples.
  • Humans: Learn through direct experiences, conversations, and exploration. Our learning encompasses everything from mastering new skills to understanding the beauty of natural phenomena.

Speed of Thinking:

  • AI: Executes tasks at lightning speed, especially those it has been trained on. It’s highly efficient with repetitive and well-defined activities.
  • Humans: May take a bit more time, but we excel at tackling complex problems and applying creative thinking to find innovative solutions.

Memory Capabilities:

  • AI: Retains information and facts based on programming but lacks personal experiences or emotions. It recalls data without the context of personal significance.
  • Humans: Remember events, emotions, and personal details, like cherished memories or favorite songs. Our memories are a blend of both joyful and challenging experiences.

Emotional Understanding:

  • AI: Does not experience emotions. It operates based on predefined rules and patterns.
  • Humans: Experience a wide range of emotions, such as joy, sadness, and empathy. Our feelings significantly influence our actions and interactions with others.

Adaptability:

  • AI: Functions based on its training and may struggle with unfamiliar situations. It is highly efficient but limited to its programmed scope.
  • Humans: Are highly adaptable and resourceful. We thrive in new and unpredictable scenarios, using creativity and critical thinking to navigate challenges.

Creativity and Innovation:

  • AI: Creates within its defined parameters. It’s like an artist working with a specific set of tools and constraints.
  • Humans: Are adept at generating novel ideas and solutions. Our creativity allows us to invent new concepts and push beyond existing boundaries.

Understanding Context:

  • AI: Knows what it has been taught but might struggle with nuanced contexts, such as interpreting humor or cultural subtleties.
  • Humans: Have a deep understanding of complex social interactions, humor, and cultural nuances. Our brains integrate diverse types of information to make sense of the world.

Decision-Making:

  • AI: Makes decisions based on data and programming. It follows logical rules and algorithms.
  • Humans: Use a combination of logic, emotions, and personal values to make decisions. Our choices reflect our unique perspectives and experiences.

In summary, AI and human intelligence each excel in their own ways. AI is a powerhouse of speed and precision for specific tasks, while human intelligence is enriched with creativity, emotional depth, and adaptability.

Summary

In this blog, we’ve explored the fundamental concepts of Artificial Intelligence (AI), starting from its basic definition to its real-world applications. We’ve simplified AI by comparing it to a robot that helps you find a lost toy, illustrating how AI mimics human-like problem-solving abilities. We discussed various everyday uses of AI, such as in fitness apps, navigation tools, and smart home devices, demonstrating its pervasive impact on our daily lives.

We also categorized different types of AI, including Narrow AI, which excels at specific tasks, General AI, which remains a theoretical concept with broader capabilities, and Artificial Superintelligence (ASI), a futuristic idea of AI surpassing human intelligence. Lastly, we introduced Generative AI, a type of Narrow AI that creates new content, showcasing its innovative potential.

This foundational understanding sets the stage for our upcoming discussions, where we’ll delve deeper into machine learning, deep learning, and the specifics of Generative AI. Stay tuned as we continue to break down these exciting topics in the next parts of our series.

How to Extract Document Information using Generative AI in 10 Minutes

How to Extract Document Information using Generative AI in 10 Minutes

The SAP Document Information Extraction Service is a solution designed to automate the extraction of data from various document types, such as invoices, purchase orders, and contracts.

It leverages machine learning and artificial intelligence to analyze and interpret document content, transforming unstructured data into structured formats that can be easily integrated into business processes.

In this detailed tutorial, we walk you through the fundamentals of SAP Document Information Extraction Service. We will show you how to configure schema, upload custom documents like invoice and resume to extract information from it.

Let’s dive in, 

You can use SAP BTP Trial Account to complete this guide. If you don’t have an SAP BTP account follow this tutorial.

Step 1: Logon to SAP BTP Trial.

Step 2: From global account click on Boosters in left pane.

Step 3: Search for “Document” and choose “Set up account for Document Information Extraction“.

Step 4: Click on Start, and choose a subaccount and click on Finish.

Step 5: On the new screen click on Go to Application. It will take you to the Document Information Service window.

Step 6: From the new screen click on highlighted icon on the left to create a schema.

Note: A schema defines how data will be extracted from documents, specifying which fields are relevant and how they should be structured. This helps ensure that the information is organized and easily accessible.

Step 7: Click on Create. 

And provide a name for schema. Choose “custom” as the Document Type. And press Create.

Note: We’re going to upload a sample invoice and do extract the content in it. In order to extract information from the invoice document we’ve to define each field in the below screens.

 

Step 8: Click on the schema that we have just created. 

Step 9: Click on ADD to add Header fields.

Step 10: Enter the fields with their respective data type that you want to extract from the document as given below.

For example: in order to extract date from the invoice provide Name as DATE, and define data type as “date”. Then, click on Save.

Step 11: After entering all the relevant header and item details click on Activate.

Step 12: Click on Document icon on the left. 

Step 13:  Add the sample invoice by clicking on the plus icon “+“. You can choose any invoice document for testing.

Step 14: Choose Document Type as “Custom”. Then, choose the schema that was created previously.

Click on plus icon “+” .

Step 15: Choose your file to test. And click on Open.

Step 16: Click on Confirm.

Step 17: Once you find document status as READY . Click on it.

Step 18: Click on Extraction Results.

Congratulations! You’ve successfully set up the SAP Document Extraction Service and extracted information from the document using advanced AI language models.

Let’s take a look at another example of document analysis.

You can create custom schemas for different types of documents. Don’t forget to select the appropriate schema when adding custom documents. Below is a snapshot of how the schema appears for a resume.

In this tutorial, we’ve examined how the SAP Document Extraction Service, combined with Generative AI and advanced extraction techniques, can revolutionize the management of custom documents. By following the steps provided, you’ve learned to streamline a previously tedious and error-prone process, allowing for quick and accurate insights from large volumes of data and complex documents.

Here is a detailed demo to help you learn further!

Build Your First SAP Gen AI Model with SAP AI Launchpad in Just 10 Minutes!

Build Your First SAP Gen AI Model with SAP AI Launchpad in Just 10 Minutes!

, , , , ,

In this blog, we’ll explore how to get started with SAP’s GenAI framework. It involves setting up of SAP AI Core and SAP AI Launchpad, and we will test the foundation model based on ChatGPT. 

Let’s dive into setting up SAP AI Core and AI Launchpad

Step 1: Provision SAP AI Core in your SAP BTP Global Account.

We’ve a free tier account(not the trial). A free tier account is needed to use SAP AI Core. 

Logon to SAP BTP Account and navigate to subaccount. My subaccount name is Zequance AI Trial. 

Get a free tier account, if you’re new to SAP BTP please watch this video.

Click on Entitlements on the left pane.

Under manage assignments click on edit to add a service plan.

Then click on Add service plans.

Search “sap ai” and choose SAP AI Core. Then, choose Free Tier and Extended plan.

Click on Add 2 service plan. 

Note: A free-tier account is required. SAP AI Core will not be available under plans if you’re using a 90-day trial account. Additionally, remember to save your changes after adding services.

Now under entitlements you will be able to see SAP AI Core as shown below.

Step 2: Select Boosters from the navigation pane on the left, then choose the booster for SAP AI Core.

Click on Start.

Once Checking is done, click on Next.

If you have a subaccount already, choose select subaccount.

Click on Next.

Click on Finish.

 

Step 3: The next step is to create keys. In the SAP BTP Cockpit, go to the Subaccounts section, select Services from the left-hand menu, and then click on Instances and Subscriptions.

To create the keys needed to access your instance, click the three dots and select Create Service Key.

Enter a service key name, for example “Booster” and click on Create.

Once your keys have been created, you can view or download them anytime by locating the key, clicking the three dots, and selecting the desired option from the available choices.

Step 4: Next, run the booster for SAP AI Launchpad. Select Boosters from the navigation pane, then choose the booster for SAP AI Launchpad from the list.

Click start.

Once checking is completed , click on Next.

Choose select subaccount and click on  Next.

Click on Next.

Click on Finish.

Step 5: Click on Instances and Subscriptions from Subaccounts.

Next, click on Go to Application. It will open the SAP AI Launchpad.

Step 6: Next configure AI API Connnection.

From the AI Launchpad screen click on Add.

We need to copy and paste the keys that we have created earlier in the step 3 to AI API Connection in Launchpad as given below. Then click on Create.

Step 7: Checking for foundation-models scenario.

Expand ML Operations in left pane and click scenarios.

There you will find foundation-models.

Step 8: Creating a configuration.

Navigate to ML Operations and then select Configurations.

  1. Provide a name for the configuration.
  2. Select the foundation-models scenario.
  3. Choose the version.
  4. Select the Executable ID.
Then click Next.

In Input Parameters, give name and version of the model you want to use. We’re using GPT models. Then, click Next.

Next, click Create.

Step 9: Create a deployment. To make LLM available for use we’ve to create a virtual LLM deployment. 

Once the configuration is complete, click on Create Deployment.

Choose Standard duration, and click on Review.

Click on Create.

After creating the deployment, wait until its status is updated to Running.

Step 10: Testing the models using Prompt Management functionality.

Navigate to Generative AI Hub>Prompt Management.

Now we cant Test the models in different approaches.

Text Summarization:

This example tasks the LLM to summarise a given text. We’re provided it with a long text and got a summary in 30 words as requested.

Question Answering:

In this test. we’ve asked the difference between SAP AI Core and the AI Launchpad. The answers we’ve got a perfecto!

Sentiment Analysis:

And finally we wanted to check if the LLM can identify sentiment of a text. 

The LLM was asked to assess the sentiment for the below text, and it was successful in identifying the “positive” sentiment mentioned in the text.

Congrats! you’ve now successfully built an SAP Foundation Model and tested it using the AI tools SAP AI Core and SAP AI Launchpad.

 

You can also view this tutorial in action on our youtube channel.

Unlock a Free 90-Day SAP BTP Trial: Explore SAP AI at No Cost!

Unlock a Free 90-Day SAP BTP Trial: Explore SAP AI at No Cost!

, , , ,

In today’s digital landscape, leveraging SAP’s Business Technology Platform (BTP) can provide your business with powerful tools for Artificial Intelligence. To get started, you’ll need to create a trial account, which allows you to explore its capabilities without any financial commitment.

Here’s a step-by-step guide on how to set up your SAP BTP trial account.

Step 1: Go to www.sap.com and click on the profile icon.

Step 2: If you do not have SAP universal ID register a new account by clicking Create your SAP account.

Then a popup window appears. Enter your details and click Submit  button.

Step 3: Next, you will receive an email from SAP asking you to verify your email address. Verify your email by clicking Click to activate your account.

Step 4: Logon to SAP BTP Trial account, https://account.hanatrial.ondemand.com/. Provide your email and click on Continue. 

Step 4: Verify your phone number by entering the code received on your phone.

Then a Legal disclaimers appears. Agree the Terms and conditions by filling check box . continue to next step by clicking on Accept.

Step 5 :  Select server which is closet to your region . By choosing AWS(prefered) click on Create Account.

Then it start creating new Account . It takes a few minutes.

Step 6 : Sometimes Org creation fails . Don’t worry we will sort it out manually later. Click on Continue.

Congrats, your trial account is ready now. Given below is your Global account view, there you see sub Accounts . Navigate to your trial sub account.

Step 7 :  Now you have landed on Sub Account.

If Org creation is successfull in step 6 , Then go to  step 10. Otherwise follow the below step.

Step 8 : To create Org manually click on cloud foundary.

Step 9 : Click on Create.

Step 10 : Explore SAP BTP Services by clicking on Service Marketplace.

Fantastic, you’ve completed the first step towards discovering SAP AI. We will show you how to use SAP AI services in the upcoming blogs.

Here's a detailed tutorial about creating SAP BTP Trial account to watch and follow!

Getting Started with SAP AI Core : A Quick Setup Guide

Getting Started with SAP AI Core : A Quick Setup Guide

, ,

SAP AI Core is a platform offered by SAP designed to support the development, deployment, and management of artificial intelligence (AI) and machine learning (ML) solutions within the SAP ecosystem. It provides a range of tools and services that help businesses integrate AI capabilities into their operations.

In this tutorial you will learn how to setup SAP AI Core and SAP AI Launchpad.

If you’re new to SAP BTP, learn how to get your trial account in this blog.

What is SAP AI Launchpad ?

SAP AI Launchpad is designed to be a user-friendly interface that simplifies the deployment and management of AI models for business users. It provides a more accessible way to interact with AI solutions without requiring deep technical knowledge.

Lets get started with the setup of SAP AI Core and Launchpad.

Step 1: Provision SAP AI Core in your SAP BTP Global Account. I’ve a free tier account(not the trial). A free tier account is needed to try SAP AI Core.

Logon to BTP Cockpit:

Navigate to subaccount. My subaccount name is Zequance AI Trial.

Click on Entitlements on the left pane.

Under manager assignments click on edit to add a service plan.

Then click on Add service plans.

Search “sap ai” and choose SAP AI Core. Then, click on Add 1 service plan. 

Note: You need to have a free tier account. If you’re in a 90 day trial account you will not find SAP AI Core under plans. 

Now under entitlements you will be able to see SAP AI Core.

 

Step 2: Choose Boosters from the navigation pane on the left. And choose the booster for SAP AI Core.

Click on start.

Keep clicking Next until you reach Finish.

Step 3: The next step is to create Keys. In the SAP BTP Cockpit, navigate to the subaccount section, select Services from the left-hand menu, and then click on Instances and Subscriptions.

To create the keys needed to access your instance, click the three dots and select Create Service Key.

Enter a service key name and click on Create.

After your keys have been created, you can view or download them anytime by locating the key, clicking the three dots, and selecting from the available options.

step 4: Provision SAP AI Launchpad in your global account

Configure your entitlement as before, but select SAP AI Launchpad

Step 5: Next, run the booster for SAP AI Launchpad. Select “Boosters” from the navigation pane, then choose the booster for SAP AI Launchpad from the list.

Click start.

Next, click on Go to Application.

Congrats, you’ve successfully setup SAP AI Core and AI Launchpad. In the next blog we will learn how to operate with SAP AI Core.