Rakuten has demonstrated how developers can accelerate their incident response workflows and cut recovery times by integrating coding agents.
As a global company spanning e-commerce, fintech, and mobile communications, Rakuten’s engineering team builds across a complex product ecosystem requiring both speed and reliability. Yusuke Kaji, General Manager of AI for Business at Rakuten, has spent the past year pushing agentic workflows into how teams plan, build, and validate software.
Rakuten’s platform engineers use tools like OpenAI’s Codex to work faster without compromising security. Across the technology sector, integrating automation into incident management drives better resilience and higher ROI.
The company’s engineers integrated Codex across operations and software delivery. This integration compressed incident response and secured an approximate 50 percent reduction in mean time to recovery (MTTR).
“We don’t just care about generating code quickly,” Kaji says. “We care about shipping safely. Speed without safety is not success.”
Coding agents accelerate incident response for Rakuten’s developers
Recovery time is a massive component of development velocity, extending beyond just shipping features. Site reliability engineering (SRE) teams often struggle with the manual workload of stitching together logs, queries, and patches during an outage.
To monitor APIs and evaluate telemetry, Rakuten teams utilise Azure’s KQL system. Codex integrates alongside these operational workflows to identify root causes and suggest remediations, shortening the delay between an alert and its resolution.
From an SRE perspective, this shortens the path from detection to remediation. By allowing engineers to focus entirely on validating and deploying fixes, this approach reduces MTTR by roughly 50 percent during outages.
In practical terms, the engineering team resolves problems twice as fast when breakages occur. This integration provides tangible efficiency improvements for complex infrastructures, offering a template for architects looking to optimise their own toolchains.
Embedding compliance into CI/CD pipelines
Accelerated shipping cycles – and the necessity for rapid incident response – often create bottlenecks during review and deployment phases. Rakuten tackles this friction by invoking Codex directly within its CI/CD pipeline. The coding agent conducts code reviews and vulnerability checks before any changes reach the production environment.
To maintain strict governance, Rakuten feeds internal coding principles into these workflows. This ensures that reviews align precisely with company expectations.
“We provide our internal coding principles to Codex,” Kaji says. “Using the same principles, it reviews whether the code aligns with our standards.”
These safety checks happen consistently and automatically, enabling teams to operate faster without lowering standards. Platform engineering leads can replicate this pattern by linking their chosen ecosystem – whether that’s OpenAI, GitHub Copilot, or GitLab Duo – directly to internal compliance repositories.
Automating complex specification builds
Beyond improving incident response, another priority for Rakuten – dubbed “AI-nisation” – focuses on enabling autonomous execution. The coding agent reads partial requirements and drives larger and more ambiguous projects forward toward working implementations without needing perfectly defined instructions.
“The latest Codex models can read between the lines,” Kaji says. “Even if the requirements are not perfectly defined, it understands what we’re trying to build.”
During one project, developers tasked Codex with building a mobile app version of an existing web-based AI agent service. The tool implemented a full-stack solution, writing a Python/FastAPI backend and a Swift/SwiftUI iOS application, alongside all necessary backend APIs. This process required no step-by-step human instruction and compressed the development timeline from an entire quarter to mere weeks.
As automation assumes more responsibility for code generation and incident response, the primary role of the software engineer transitions toward writing clearer specifications and verifying outputs against defined standards. To adapt to this new environment, technical architects must prioritise system design over syntax.
“Our role is not to check every line of code anymore,” Kaji says. “Our role is to define clearly what we want and establish how to verify it.”
Rakuten supports this evolution by hosting hands-on workshops across engineering, product, and non-technical teams. This contributes to Codex playing a central role in helping teams ship faster, operate more safely, and scale autonomous development across the organisation.
For teams building complex applications, the advice is to invest heavily in test-driven development and prompt engineering training. By establishing exact verification methods upfront, teams can safely delegate the heavy lifting to autonomous tools like coding agents which help to scale development and incident response workflows efficiently across the organisation.
See also: BMC: Integrating mainframe systems into modern CI/CD pipelines
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.