9 hours ago
Rakuten is using Anthropic’s AI coding agent, Claude Code, to transform software development, enabling their engineering teams to automate coding tasks that accelerate time to market while reliably powering AI features for millions of customers.
Key results with Claude Code
Achieved 7 hours of sustained autonomous coding on a complex open-source refactoring project
Reduced time to market for new features by 79% (from 24 days to 5 days)
Delivered 99.9% accuracy on complex code modifications
Empowering every team to innovate at speed
With over 70 businesses spanning e-commerce, travel, fintech, digital content and communications, Rakuten is using AI to better serve their customers, business partners and employees. “We want to give all our teams the power to innovate quickly and drive greater impact for customers faster,” said Yusuke Kaji, General Manager of AI for Business at Rakuten. “It’s about multiplying what each team can achieve, not just automating existing tasks.”
The scale of Rakuten’s ambition presented unique challenges. With thousands of developers serving millions of customers, they needed AI to accelerate time to market while maintaining enterprise-grade quality and security standards. “Speed and ROI are our key metrics,” Kaji explained. “We measure success by how quickly we can deliver value to customers, not just how fast AI can write code.”
Rakuten is weaving AI throughout its entire business operations as part of its “AI-nization” strategy. This includes building AI agents and large language models from the ground up, giving the company a unique understanding of the technology’s potential. That foundational approach, coupled with a commitment to providing Rakuten employees with the best tools available, inspired the team to push the boundaries of AI-assisted software development.
Selecting Claude Code for enterprise-scale development
Rakuten’s evaluation of existing AI coding tools revealed a consistent problem: most required constant human guidance and couldn’t navigate complex, multi-language codebases. Claude Code solved both issues, delivering genuine autonomous development at enterprise scale.
This autonomy has started to multiply engineering capacity across Rakuten. “You can have five tasks running in parallel by delegating four to Claude Code while focusing on the remaining one,” Kaji explained.
An inflection point came when Kenta Naruse, Machine Learning Engineer, tested Claude Code’s capabilities by giving it a complex technical task: implement a specific activation vector extraction method in vLLM, a massive open-source library with 12.5 million lines of code in multiple programming languages. Claude Code finished the entire job in seven hours of autonomous work in a single run. “I didn’t write any code during those seven hours,” Naruse recalled. “I just provided occasional guidance.” The implementation achieved 99.9% numerical accuracy compared to the reference method, reinforcing that Claude Code could handle Rakuten’s advanced engineering challenges.
This breakthrough, combined with Anthropic’s commitment to responsible AI that aligned with Rakuten’s values, convinced leadership that Claude Code could transform their development process.
Transforming development workflows across the organization
Instead of retrofitting existing processes with AI, Rakuten redesigned their development workflows around Claude Code’s capabilities.
Engineers use Claude Code throughout the development lifecycle: writing unit tests, mocking APIs, building components, fixing bugs and generating documentation. New team members leverage Claude to understand complex codebases and architectural decisions, dramatically accelerating their ability to contribute meaningfully to projects.
This accelerated development directly benefits Rakuten’s organization and customers alike. Teams ship features in days instead of weeks with augmented workflows:
AI-powered code review: Provides instant feedback on pull requests, improving code quality and development speed
Parallel development: Runs multiple Claude Code sessions simultaneously to eliminate bottlenecks and accelerate feature delivery
The transformation extends to how individual developers work. Diego Mateos, Senior Machine Learning Engineer at Rakuten, discovered Claude Code changed his approach to coding, saying “I wasn’t naturally using test-driven development before, but Claude Code makes it so easy. It generates comprehensive tests instantly, then builds features that pass them. It’s completely changed how I develop and made me a much more efficient engineer.”
Achieving measurable business impact
The results have been impactful for both Rakuten’s business and their employees. Engineering teams saw their average time to market for new features drop from 24 working days to 5 days—a 79% reduction that changed how Rakuten delivers innovation. “Our time to market has been significantly brought down because Claude Code gives us those super powers to make executions much, much faster,” said Manoj Desai, Manager at Rakuten’s AI Empowerment Section. The seven-hour autonomous coding session became a turning point for Rakuten. “After the announcement, teams from across the company were asking about it,” Naruse shared. “When I explained that Claude Code did it with minimal human input, people were stunned. They realized this could transform how we build software.”
Beyond speed, Claude Code has democratized development capabilities across Rakuten. “We asked non-engineers to use Claude Code as well,” said Kaji. “Thanks to the terminal interface, colleagues from non-technical roles can contribute to coding projects without directly editing code. With appropriate context and coding guidelines, Claude Code acts as a safety guardrail.” This expansion of contributors to technical projects has accelerated innovation velocity, one of Rakuten’s key metrics alongside test coverage and “AI-nization ratio.”
The success has inspired even more ambitious projects. Naruse is now building an “ambient agent” that breaks complex tasks into 24 parallel Claude Code sessions, each handling different aspects of updating Rakuten’s massive monorepo. This project would typically take over a month to complete manually.
Rakuten and Anthropic: Building the future of AI-powered development
Following successful pilot implementations, Rakuten’s leadership is committed to scaling Claude Code access across their massive developer organization. The company is expanding Claude Code’s capabilities from handling quick tasks to managing complex, multi-hour workflows that will become routine parts of development.
“Claude Code is not just a tool—it’s part of our AI journey,” said Desai. This journey, built on shared values between Rakuten and Anthropic around developing helpful, honest and harmless AI, aims to responsibly transform how software is created and delivered to customers.
Rakuten’s AI-nization offers a blueprint for enterprise transformation. By giving thousands of employees access to AI that can code autonomously and significantly augment engineers’ capabilities, Rakuten is creating an organization where technical barriers no longer limit innovation. Complex technical challenges become manageable, non-engineers can contribute to development and customer features ship in days instead of months.
https://rakuten.today/blog/...elopment-with-claude-code%ef%bf%bc.html
|
Thread abonnieren
