The relentless pressure to deliver software faster whereas maintaining stability is a constant challenge for development teams. Increasingly, companies are turning to artificial intelligence, specifically coding agents, to accelerate incident response and reduce downtime. Rakuten, a global technology company with interests spanning e-commerce, fintech, and mobile communications, has demonstrated significant success in this area, achieving a roughly 50 percent reduction in signify time to recovery (MTTR) by integrating tools like OpenAI’s Codex into its workflows.
Rakuten’s experience highlights a growing trend within the technology sector: the automation of incident management. For organizations managing complex product ecosystems, the ability to quickly identify and resolve issues is paramount. Yusuke Kaji, General Manager of AI for Business at Rakuten, has been instrumental in driving the adoption of agentic workflows across the company’s engineering teams, focusing on how AI can enhance planning, building, and validation processes. This isn’t simply about speed, Kaji emphasizes; it’s about “shipping safely,” recognizing that rapid deployment without robust safeguards can introduce new vulnerabilities and risks.
The core of Rakuten’s approach involves leveraging coding agents – AI-powered tools capable of understanding and generating code – to streamline the incident response process. These agents aren’t intended to replace engineers, but rather to augment their capabilities, freeing them from repetitive tasks and allowing them to focus on more complex problem-solving. This integration extends beyond simply generating code; it encompasses code review, vulnerability checks, and even the automated creation of fixes, all within the existing CI/CD pipeline.
Accelerating Incident Response with AI-Powered Tools
Traditional incident response often involves a manual and time-consuming process of sifting through logs, analyzing telemetry data, and crafting patches. Site Reliability Engineering (SRE) teams, responsible for ensuring the reliability of systems, frequently find themselves bogged down in these tasks. Rakuten’s implementation addresses this bottleneck by integrating coding agents with existing operational workflows, such as Azure’s KQL system for API monitoring and telemetry evaluation. By automating the identification of root causes and suggesting potential remediations, the company has significantly shortened the time between an alert being triggered and a resolution being implemented.
The impact of this integration is substantial. Rakuten engineers are now resolving issues twice as quickly when problems arise, representing a tangible improvement in efficiency for complex infrastructures. This success story provides a blueprint for other organizations looking to optimize their toolchains and enhance their resilience. The key, according to Rakuten, is to prioritize safety alongside speed, ensuring that automated processes adhere to strict governance and compliance standards.
Embedding Compliance into the Development Lifecycle
One of the challenges of accelerating software delivery is maintaining code quality and adhering to internal coding standards. Rakuten tackles this issue by invoking coding agents directly within its Continuous Integration/Continuous Delivery (CI/CD) pipeline. Before any code reaches the production environment, the coding agent conducts automated code reviews and vulnerability checks. This proactive approach helps to identify and address potential issues early in the development process, reducing the risk of introducing bugs or security flaws.
To ensure consistency and alignment with company expectations, Rakuten feeds its internal coding principles directly into the coding agent’s workflows. This allows the agent to evaluate code against established standards, providing a consistent and objective assessment. Platform engineering teams can replicate this pattern by connecting their chosen AI ecosystem – whether it’s OpenAI, GitHub Copilot, or GitLab Duo – to internal compliance repositories. This ensures that all code changes are automatically vetted against the organization’s specific requirements.
Automating Complex Specification Builds and the Future of Software Engineering
Rakuten’s ambitions extend beyond simply improving incident response. The company is also exploring how coding agents can enable autonomous execution, allowing them to drive larger and more ambiguous projects forward with minimal human intervention. This involves leveraging the ability of advanced models, like Codex, to “read between the lines” and understand project requirements even when they are not perfectly defined.
A compelling example of this capability involved the development of a mobile app version of an existing web-based AI agent service. Rakuten developers tasked Codex with building the application, and the agent autonomously delivered a full-stack solution, including a Python/FastAPI backend and a Swift/SwiftUI iOS application, complete with all necessary backend APIs. Remarkably, this process was completed in a matter of weeks, a timeline that would have previously taken an entire quarter.
As automation takes on a greater role in code generation and incident response, the role of the software engineer is evolving. The focus is shifting from writing code to defining clear specifications and verifying the outputs generated by AI agents. Technical architects must prioritize system design over syntax, ensuring that the overall architecture is robust and scalable. According to Kaji, “Our role is not to check every line of code anymore. Our role is to define clearly what we desire and establish how to verify it.”
Rakuten is actively supporting this transition by providing hands-on workshops for engineering, product, and non-technical teams. These workshops aim to equip employees with the skills and knowledge needed to effectively collaborate with AI agents and leverage their capabilities. Investing in test-driven development and prompt engineering training is also crucial, as it enables teams to establish clear verification methods and safely delegate tasks to autonomous tools.
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security & Cloud Expo. Click here for more information.
Developer is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.
The integration of coding agents into development workflows represents a significant step towards a more automated and efficient software development process. As AI technology continues to evolve, we can expect to witness even more sophisticated applications emerge, further transforming the role of the software engineer and accelerating the pace of innovation. The ongoing development and refinement of these tools, and the strategies for their responsible implementation, will be key to unlocking their full potential.
Rakuten’s continued investment in “AI-nisation” and its commitment to fostering a culture of experimentation suggest that this is just the beginning of a broader transformation within the company and potentially across the industry. The next step for Rakuten will be to further refine its agentic workflows and explore new use cases for AI-powered automation, with a continued focus on safety, reliability, and scalability.
Key Takeaways:
- Coding agents can significantly reduce mean time to recovery (MTTR) by automating incident response tasks.
- Integrating AI into the CI/CD pipeline enhances code quality and compliance.
- The role of the software engineer is evolving towards specification writing and output verification.
- Investing in test-driven development and prompt engineering is crucial for successful AI integration.
What are your thoughts on the role of AI in software development? Share your comments below, and let’s continue the conversation.
Worth a look