Google Vertex AI

  • Author: Ronald Fung

  • Creation Date: 14 June 2023

  • Next Modified Date: 14 June 2024

A. Introduction

Vertex AI is a machine learning (ML) platform that lets you train and deploy ML models and AI applications. Vertex AI combines data engineering, data science, and ML engineering workflows, enabling your teams to collaborate using a common toolset.

Vertex AI provides several options for model training:

  • AutoML lets you train tabular, image, text, or video data without writing code or preparing data splits.

  • Custom training gives you complete control over the training process, including using your preferred ML framework, writing your own training code, and choosing hyperparameter tuning options.

After you deploy your models, use Vertex AI’s end-to-end MLOps tools to automate and scale projects throughout the ML lifecycle. These MLOps tools are run on fully-managed infrastructure that you can customize based on your performance and budget needs.

You can use the Vertex AI SDK for Python to run the entire machine learning workflow in Vertex AI Workbench, a Jupyter notebook-based development environment. Other available interfaces include the Google Cloud Console, the gcloud command line tool, client libraries, and Terraform (limited support).

B. How is it used at Seagen

Google Vertex AI is a suite of machine learning tools and services that enable businesses to build, train, and deploy machine learning models on Google Cloud. Here are some ways that Seagen can use Google Vertex AI to improve their machine learning processes:

  1. Data Preparation: Google Vertex AI provides tools for data preparation and cleaning, enabling businesses to prepare their data for machine learning tasks. Seagen can use Google Vertex AI to clean and preprocess their data, ensuring that it is ready for machine learning tasks.

  2. Model Training: Google Vertex AI provides tools for model training, enabling businesses to train machine learning models quickly and efficiently. Seagen can use Google Vertex AI to train their machine learning models and ensure that they are accurate and effective.

  3. Model Selection: Google Vertex AI provides tools for model selection, enabling businesses to select the best machine learning model for their specific use case. Seagen can use Google Vertex AI to evaluate and compare different machine learning models and select the best one for their needs.

  4. Model Deployment: Google Vertex AI provides tools for model deployment, enabling businesses to deploy their machine learning models in production environments quickly and easily. Seagen can use Google Vertex AI to deploy their machine learning models and ensure that they are performing as expected.

  5. AutoML: Google Vertex AI provides AutoML tools that enable businesses to automatically build, train, and deploy machine learning models without requiring extensive machine learning expertise. Seagen can use Google Vertex AI AutoML to build and deploy machine learning models quickly and efficiently, without the need for extensive machine learning expertise.

By using Google Vertex AI, Seagen can improve their machine learning processes, accelerate time-to-market for their products, and reduce the risk of errors and downtime. Google Vertex AI provides a range of tools and services that enable businesses to build, train, and deploy machine learning models on Google Cloud, making it a powerful solution for machine learning tasks.

C. Features

Choose a model that fits your needs

Jumpstart your ML project with Model Garden, a single place to access a wide variety of APIs, foundation models, and open source models. Kick off a variety of workflows including using models directly, tuning models in Generative AI Studio, or deploying models to a data science notebook.

End-to-end MLOps

Vertex AI provides purpose built tools for data scientists and ML engineers to efficiently and responsibly automate, standardize, and manage ML projects throughout the entire development life cycle. Using Vertex AI you can easily train, test, monitor, deploy, and govern ML models at scale, reducing the work needed to maintain model performance in production and enabling data scientists and ML engineers to focus on innovation code.

Data and AI integration

Through Vertex AI Workbench, Vertex AI is natively integrated with BigQuery, Dataproc, and Spark. You can use BigQuery ML to create and execute machine learning models in BigQuery using standard SQL queries on existing business intelligence tools and spreadsheets, or you can export datasets from BigQuery directly into Vertex AI Workbench and run your models from there. Use Vertex Data Labeling to generate highly accurate labels for your data collection.

Low-code and no-code tooling

Vertex AI provides low-code tooling and up-training capabilities so practitioners with a wide variety of expertise can leverage machine learning workloads. With Generative AI Studio, developers can tune and deploy foundation models for their use cases via a simple UI. And, with our off the shelf APIs, developers can easily call upon pre-trained models to quickly solve real-world problems.

Open and flexible AI infrastructure

Vertex makes it easy to deploy ML models to make predictions (also known as inference) at the best price-performance for any use case. It provides a broad selection of ML infrastructure and model deployment options to help meet all your ML inference needs. It is a fully managed service and integrates with MLOps tools, so you can scale your model deployment, reduce inference costs, manage models more effectively in production, and reduce operational burden.

D. Where Implemented

LeanIX

E. How it is tested

Testing Google Vertex AI involves verifying that the machine learning models are accurate, effective, and performing as expected. Here are some steps you can take to test Google Vertex AI:

  1. Define Machine Learning Tasks: Define the machine learning tasks for your application, such as image classification or natural language processing. Create the necessary data sets and configurations needed for the machine learning model to be built and trained.

  2. Build Machine Learning Models: Build the machine learning models using the necessary tools and verify that the build process is successful. This can include checking that all data sets are included, that the model is built correctly, and that any tests have passed.

  3. Train Machine Learning Models: Train the machine learning models using the necessary tools and verify that the training process is successful. This can include checking that the model is accurate and effective and that any training metrics meet the desired thresholds.

  4. Deploy Machine Learning Models: Deploy the machine learning models to your target environment, such as a Google Cloud project, and verify that it works correctly. This can include testing that the model is accessible, that it performs as expected, and that there are no errors or bugs.

  5. Monitor Performance: Monitor the performance of the machine learning models to ensure that they are stable, accurate, and performing as expected. This can include monitoring model accuracy, performance metrics, and other model-specific metrics.

  6. Perform A/B Testing: Perform A/B testing to compare the performance of different machine learning models and ensure that they meet the desired performance thresholds.

By following these steps, you can test Google Vertex AI and ensure that your machine learning model building, training, and deployment process is working correctly and producing the expected results. It’s important to regularly test your machine learning models to ensure that they remain accurate, effective, and reliable and to avoid costly errors or downtime.

F. 2023 Roadmap

????

G. 2024 Roadmap

????

H. Known Issues

While Google Vertex AI is a powerful suite of machine learning tools and services, there are some known issues or limitations that you should be aware of. Here are some of the common issues with Google Vertex AI:

  1. Limited Customization: While Google Vertex AI provides tools for machine learning model building, training, and deployment, the customization options may be limited for more complex machine learning models. It’s important to review the customization options and ensure that they meet the specific needs of your application.

  2. Cost: Google Vertex AI is a paid service, and the cost can increase significantly for businesses with large or complex machine learning models. It’s important to review the pricing structure and estimate the cost of the service for your specific needs.

  3. Limited Integration with Non-Google Services: While Google Vertex AI integrates well with other Google services, it may not integrate with all third-party services or tools. It’s important to review the integration options and ensure that all necessary integrations are supported.

  4. Limited Support for Certain Machine Learning Models: While Google Vertex AI supports a wide range of machine learning models, it may not support all models or algorithms. It’s important to review the supported models and ensure that they meet the specific needs of your application.

  5. Network Latency: Google Vertex AI may experience network latency, especially for businesses that are located far from the Google Cloud data centers. It’s important to monitor network latency and adjust the storage settings as needed.

It’s important to be aware of these limitations and issues when using Google Vertex AI. By understanding these challenges, you can better ensure that your machine learning processes remain reliable and efficient and avoid costly errors or downtime.

[x] Reviewed by Enterprise Architecture

[x] Reviewed by Application Development

[x] Reviewed by Data Architecture