[Session Report] How to Create the Best Software Development Environment with Google Cloud + GitLab #GoogleCloudNextTokyo

I participated in Google Cloud Next Tokyo 2025 held on August 5-6, 2025!
In this article, I'll report on one of the sessions titled "How to Create the Best Software Development Environment with Google Cloud + GitLab."

https://www.googlecloudevents.com/next-tokyo/

Session Overview

Title

"How to Create the Best Software Development Environment with Google Cloud + GitLab"

Summary

In a landscape crowded with generative AI, MCP servers, and various source code generation technologies, there are still applications that need to be written by humans. When considering Google Cloud as a deployment destination, what does the best software development environment that fully leverages Google Cloud/Gemini technologies look like? No matter what technology is used to write the system, everything can work out as long as there's a mechanism to prevent regressions. Please join this high-energy session to see how to create a comfortable and optimal software maintenance and development system.

Speaker

GitLab G.K.
Senior Solution Architect
Tsukasa Komatsuhara

Main Google Cloud products/services covered

・Cloud Run
・Cloud Workstations
・Google Kubernetes Engine (GKE)

https://www.googlecloudevents.com/next-tokyo/session/3123557/google-cloud-+-gitlabで作る最高のソフトウェア開発環境の作り方

About GitLab G.K.

GitLab was incorporated in 2014 and is a company where all employees work remotely.
It is used by over 100,000 organizations and more than 30 million users worldwide, and their knowledge is summarized in the book "The Remote Playbook."
The company has been recognized as a "Leader" in the 2024 Gartner® Magic Quadrant™ for DevOps Platforms and has received the Google Cloud Technology Partner of the Year award for multiple years, making it a leading presence in the industry.### The Serious Damage Caused by "Uncomfortable" Environments

• Waste of valuable time

  • Inefficient onboarding
    Every time a new member joins the team, existing members spend time setting up their PC. While procedure manuals exist, they often contain outdated or incorrect information, resulting in the need for one-on-one intensive support, which takes away from actual work time.

• Review process
  With the proliferation of generative AI, the volume of code requiring review has increased explosively, pushing reviewers to their limits. As a result, to avoid the enormous workload of reviews, code is often approved with insufficient checks—"just going through the motions"—leading to the quality assurance process becoming merely a formality.

• Neglected security risks
  Even when critical vulnerabilities are discovered late in the development process, such as just before release, schedule adherence often takes priority, and products may be released without fixes. Temporary measures like firewall protections instead of fundamental solutions lead to the serious problem of steadily accumulating technical debt.

As a solution to these issues, we presented "The Ideal State of Future Software Development Environments."

1. Environment: Always being able to prepare comfortable development environments.
2. Review: Ensuring both humans and AI review not only code but also "working environments" as a set.
3. Safety and quality: Making security scans mandatory and steadily resolving vulnerabilities with AI support.
4. Innovation: Being able to incorporate evolving source code generation AI technology while maintaining control.

And as concrete methods to realize these, an advanced workflow combining Google Cloud and GitLab was introduced.

Building a Comfortable and Secure Development Foundation with Cloud Workstations

First, we unify the entry point for development environments with "Cloud Workstations" and GitLab.

Mechanism
Developers use highly functional development environments like VS Code in their browsers. Source code is managed in GitLab and never stored on local PCs.

Benefits

• Prevention of information leaks
  Prevents source code from leaking externally, enabling secure development.
• Immediate deployment and consistency of environments
  By "imaging" the entire development environment, issues with dependencies such as Python version differences can be avoided. New members can instantly obtain the same environment as everyone else with just one button.
• Increased productivity
  Building, testing, and even container creation can be performed seamlessly within the environment.##### Improving Quality and Teamwork with AI Reviews and Review Apps

Next, let's look at the review process, which is key to quality.

Achieving "Non-offensive Reviews"
Rather than having code directly criticized by another person, using AI (GitLab Duo) as the reviewer makes it easier for developers to objectively accept feedback.

Dynamic Testing with "Review App"
GitLab pipelines automatically run builds and tests every time code is pushed.
After container scanning, SAST (Static Application Security Testing), secret detection, SBOM (Software Bill of Materials) generation, etc., temporary review applications (Review Apps) are automatically deployed on Cloud Run or GKE.
Reviewers can check not only the code but also the actual working application, which significantly improves the accuracy of reviews.

Democratizing Vulnerability Response with AI

Security scanning is essential, but fixing discovered vulnerabilities used to require specialized knowledge.
In GitLab, AI supports this process.

• AI-based Solutions and Explanations
  In vulnerability reports, you can request AI to "explain" or "resolve" issues.

• Automatic Fixes
  By selecting "resolve," GitLab Duo AI automatically generates a merge request with proposed fixes. This enables vulnerability response even for those who aren't security specialists.

Future Outlook: AI as a Quality Gateway

Finally, Mr. Komatsubara discussed his vision of using GitLab as a "quality gate" in a future where various AI code editors become prevalent.
"This is an advanced concept where platform AI (GitLab Duo) evaluates and controls the quality of code generated by various AI tools. AI will summarize code reviews, troubleshoot CI/CD issues, and even provide dashboards showing the effects of AI implementation, supporting the entire development process."

Impressions

This session clearly showed that investing in development environments is no longer just a cost, but a strategic move to unleash developer productivity and creativity, directly enhancing business competitiveness. The solutions provided by Google Cloud and GitLab are powerful weapons for modern development teams to solve the challenges they face and deliver higher quality software to the market more quickly.