Classmethod accelerates development velocity with Claude Code
.png){kind=logo}
Industry:
Software
Company size:
Large
Product:
Claude Code
Location:
Asia Pacific
80% decrease
in time spent reviewing code
96% reduction
in Google Apps Script task processing time, from 24 hours to 1 hour
The problem
The solution
Japanese cloud integrator Classmethod empowers engineering teams to overcome chronic labor shortages by achieving up to 10x productivity gains through AI-driven development with Claude Code.
With Claude Code, Classmethod achieved:
- Reduced development time for specific coding tasks by up to 90%
- Up to 80% decrease in time spent reviewing code
- 96% reduction in Google Apps Script task processing time, from 24 hours to 1 hour
- Increase in code contributions shipped to production (108 merged pull requests in January to 165 in May-June), enabling the team to deliver features and fixes faster
Efficient utilization of limited engineering resources
Japan's technology sector is navigating an evolving landscape where demand for skilled engineering talent continues to grow. Development teams are exploring innovative approaches to optimize their workflows and meet expanding business needs. At Classmethod, the team recognized an opportunity to enhance how engineers allocate their time, enabling them to focus more on strategic and creative initiatives alongside their core development work.
"Engineers were unable to dedicate sufficient time to the tasks they should be focusing on." The gap between available developer hours and project requirements continued to widen, threatening their ability to deliver on client expectations while maintaining code quality.
Selecting Claude Code for superior code generation
After evaluating multiple AI agentic coding solutions, Classmethod chose Claude Code because "the quality of the generated code was significantly superior to that of other products." During an initial pilot, Claude Code demonstrated exceptional accuracy and faithfulness to instructions, generating practical code optimized for their specific technology stack.
The tool's design as a command-line interface meant virtually zero learning curve for developers already comfortable with terminal workflows. "It has a low learning curve as a CLI tool," they said, noting seamless integration with existing development environments.
Transforming the Development Lifecycle Across Various Projects
Classmethod integrated Claude Code across their coding workflows, from initial implementation to final review.
The tool generates entire logic blocks from detailed designs, helps developers understand existing codebases 90% faster, and creates comprehensive test suites. "Claude Code handles GitHub ISSUES-based implementation and test descriptions," they said, supporting their full technology stack including Ruby on Rails, Next.js, and Terraform.
Their innovation extends beyond traditional coding. Mandatory AI-powered self-reviews before pull requests are being gradually implemented in several projects, with Claude Code detecting security risks and violations of DRY and YAGNI principles—catching redundant code and unnecessary complexity before they reach production. The /explain-pr and /review-pr commands provide instant context and feedback, elevating code quality across the organization.
Achieving multi-service development with small teams
The transformation at Classmethod goes beyond raw productivity metrics.
"Parallelization of thinking enables simultaneous progress on multiple projects," they share, achieving development structures previously impossible with their team size. Engineers now handle multiple services concurrently, maintaining focus while managing numerous tasks.
When Classmethod developed "rulesync," an open-source project for their development workflows, 99% of the codebase was implemented using Claude Code. Long-standing backlog items like rich admin dashboards, Grafana MCP chat applications, and dark mode implementations finally became reality.
Google Apps Script automation particularly showcases their dramatic efficiency gains. Tasks that previously consumed 24 hours are now completed in just one hour. "The implementation accuracy is particularly excellent in GAS," they note, highlighting how Claude Code excels at understanding and generating code across frameworks and environments.
Pioneering the future of Japan's AI-driven development
Claude Code has enabled Classmethod to drive productivity gains internally and for their clients. In the future, they intend to expand their use cases, including development-adjacent tasks like library updates and detailed design generation, while exploring sub-agents functionality for even greater automation.
They plan to include infrastructure integration for automatic Cloud Run Functions upgrades in technical expansion, alongside improved UI implementation that adheres to design systems. They're also working to extend these capabilities beyond their own engineering team, democratizing access to AI-powered development tools among their clients.
By sharing AI-driven development know-how and deployment strategies, they're positioning themselves to help other Japanese companies navigate the transition to AI-augmented engineering.
Their vision is clear: transform how software gets built in Japan, enabling small teams to match the productivity of much larger organizations with Claude Code.
Video caption
Video caption
Video caption