Google Cloud AI Services

Google Cloud provides a set of AI tools and services for developers to build, train, and deploy ML models at scale. Google Cloud AI services provide the ability to build custom machine learning (ML) models using Vertex AI, which is a fully managed machine learning platform that allows developers and data scientists to build, train, and deploy machine learning models from end to end. Google Cloud AI also offers services based on pre-trained models and APIs for common ML tasks, such as image and speech recognition, natural language processing (NLP), and other predictive analytics. Google Cloud AI services integrate seamlessly with other GCP services and tools for data preprocessing, training, monitoring, and evaluation. In this chapter, we will cover the following topics:

• Google Vertex AI: This is a suite of services for users to develop ML models, including data collection, labeling, training, deploying automated machine learning (AutoML), and other...

Google Cloud Vertex AI

Vertex AI is an integrated set of products, features, and a management interface that simplifies the management of Google Cloud ML services. Vertex AI lets users build, train, and deploy ML models. As shown in Figure 10.1, Vertex AI unifies a set of disparate features and has a user interface that makes it easy to develop and integrate ML-related applications:

Figure 10.1 – Google Vertex AI suite

In this section, we will briefly discuss the following Vertex AI concepts first and then spotlight Vertex AI AutoML with a lab to train a simple ML model:

• Vertex AI datasets
• Vertex AI dataset labeling
• Vertex AI Feature Store
• Vertex AI Workbench and notebooks
• Vertex AI custom models
• Vertex Explainable AI
• Vertex AI prediction
• Vertex AI model monitoring
• Vertex AI Pipelines
• Vertex AI TensorBoard
• Vertex AI Metadata
• Vertex AI AutoML

Let us start looking at Vertex AI by looking at...

Google Cloud ML APIs

The Google Cloud ML APIs are a set of ML API services that provides pre-trained models and tools for developers and data scientists to build custom ML models and ML applications. Figure 10.8 shows a diagram of the Cloud ML APIs that we will cover next:

Figure 10.8 – Google Cloud ML APIs

Sight API

Google Cloud Sight API includes Cloud Vision API and Cloud Video API. The former analyzes static images and the latter learns from videos.

Cloud Vision API

Google Cloud Vision API is a cloud-based image analysis tool that helps developers to build applications that can understand the content of images, leveraging pre-trained computer vision models. Here are some key features of Cloud Vision API:

• Label detection: It helps to identify objects and the context within an image
• Optical Character Recognition (OCR): It detects and extracts text from images
• Face detection: It detects faces and facial features within...

Google Cloud generative AI services

Google Cloud generative AI services use large language models (LLMs) to recognize, predict, and generate human languages. They can be used in many business applications to drive new business opportunities. Figure 10.10 lists some generative AI use cases in Google Workspace applications.

Figure 10.10 – Generative AI use cases in Google Workspace

Since the needs of enterprise customers are different from individual consumers and enthusiasts, Google Cloud generative AI focuses on enterprise business cases and environments. At this time, it offers Generative AI Studio, which includes three tabs: Overview, Language, and Speech, as shown in Figure 10.11.

Figure 10.11 – Generative AI Studio in Vertex AI

In the Language tab, it provides starting points for prompts, conversations, and model tuning, as shown in Figure 10.12:

Figure 10.12 – Generative...

Summary

In this chapter, we discussed the two main topics in Google Cloud ML: Google Cloud Vertex AI, which provides an end-to-end suite for developing and deploying ML models, and Google ML API, which provides programming interfaces by leveraging pre-trained ML models. We have also covered Google generative AI, which is an emerging AI service added to Vertex AI recently.

In this chapter, you have acquired ML model training and ML application development skills using the Google Vertex AI suite and Google ML API. In the next chapter, we will discuss Google Cloud security.

Practice questions

Questions 1-4 are based on Figure 10.15, which is a Cloud Vision system to detect unsafe content in input images.

Figure 10.15 – Cloud Vision detection system

We have created three files:

• target.json
• main.py
• requirements.txt

And the following is a snippet of the main.py file:

def image_checking(data, context):
    uri = "gs://" + data['bucket'] + "/" + data['name']
    image = vision.Image()
    image.source.image_uri = uri
    response = vision_client.safe_search_detection(image=image)
    result = response.safe_search_annotation

1. What is the requirements.txt file?

A. It includes the required modules to import

B. It can be renamed to requirements

C. It is only for illustration purposes

D. It is needed for GCP Cloud Vision

2. What is the target...

Answers to the practice questions

1. A
2. A
3. A
4. A
5. A
6. A
7. A
8. D

Further reading

For further insights into what you’ve learned in this chapter, refer to the following links:

• https://cloud.google.com/products/ai
• https://cloud.google.com/vertex-ai-workbench
• https://cloud.google.com/automl
• https://cloud.google.com/vertex-ai
• https://cloud.google.com/vision
• https://cloud.google.com/video-intelligence
• https://cloud.google.com/natural-language
• https://cloud.google.com/translate