New Relic Adopted by airCloset, Reducing Problem Investigation Time by 1/20th and Saving Stylists 2,520 Hours Annually in Product Selection

New Relic, Inc. (Headquarters: Chuo-ku, Tokyo, hereinafter "New Relic"), a provider of observability platforms for digital businesses, announced today that airCloset, Inc. (Headquarters: Minato-ku, Tokyo, hereinafter "airCloset"), operator of the subscription-based fashion rental service "airCloset" for women, has implemented the observability platform "New Relic." This implementation automates the performance monitoring, improvement, and failure investigation of their core systems, building a foundation for fully autonomous operation driven by AI. As a result, problem investigation time has been reduced to 1/20th of the previous time, and product selection time for stylists has been reduced by 2,520 hours annually.

Background and Circumstances of Implementation

airCloset is a pioneer in subscription services, offering rental clothing selected by professional stylists. They operate "airCloset," boasting approximately 1.5 million members as of the end of March 2026. The company also offers a diverse range of services, including the appliance rental service "airCloset Mall," the pay-per-use fashion rental service for occasion items "airCloset Spot RENTAL," the body type and color analysis salon "airCloset Salon," and plans to launch the subscription-based men's fashion rental service "airCloset Men's" in May 2026.

The company develops its core systems in-house, supporting both its services and logistics. Approximately 70 development projects are concurrently underway. They have built a proprietary platform where web systems, styling systems, and warehouse management systems interact, ensuring speed in system improvements. However, as the scale and complexity of the systems grew, grasping the overall status became difficult, leading to challenges such as "delayed response to failures" and "difficulty in identifying performance degradation."

Despite introducing observability products to address these issues, adoption by the field was slow, and there were instances where customers noticed errors or response degradation before the company did. Particularly during peak times, database load increased, and issues like the "N+1 problem" in code caused exponentially growing load.

Implementation and Effects of New Relic

To fundamentally improve this situation, the company decided to implement New Relic in January 2023. The deciding factors for adopting New Relic were the robust provision of know-how, such as post-implementation setup support and onboarding support, and its function as an integrated platform capable of comprehensively collecting, analyzing, and visualizing the data necessary for observability, including application performance, infrastructure metrics, and operational logs.

Following the implementation of New Relic, the following effects were achieved:

Development of "New Relic Analyzer," an AIOps mechanism, reduced problem investigation time by 1/20th and eliminated man-hours from monitoring to improvement plan formulation.

airCloset developed "New Relic Analyzer," an AIOps mechanism that integrates "New Relic MCP Server" with multi-agents, automating problem investigation with AI. This mechanism was brought to a practical level within days of its design. Every morning, the AI (Gemini Agent) analyzes the status of applications monitored by New Relic MCP Server and automatically detects six types of issues, including N+1 queries and slow transactions. When a problem is detected, the source code is investigated through the in-house "Git Server MCP" to identify the problematic code, and a Backlog ticket is automatically created in both Japanese and English with a proposed fix. This realization of a state where "engineers receive improvement tickets with the problematic code upon arrival at work" has reduced problem investigation time to 1/20th of the previous time and eliminated man-hours from performance monitoring to improvement plan formulation. Furthermore, by setting call limits and time constraints for the AI, the operational scope is clearly defined, improving the completion rate of investigations.

Table: 6 Problem Types Detected by New Relic Analyzer and Detection Conditions

Source: airCloset, Inc.

Reduced Stylist's Clothing Selection Time by 2,520 Hours Annually

Through the investigation and analysis of performance issues via New Relic transactions, the time taken for stylists to select clothing suitable for customers has been reduced by 2,520 hours annually.

Achieved Autonomous Performance Improvement Cycle with Minimal AI Agent Operating Costs of Approximately $7 per Month

By proactively discovering and fixing the root causes of increased load, server load during peak times has been significantly reduced, leading to substantial infrastructure cost savings annually. Additionally, the operating cost of the AI agent is kept very low, at approximately $7 per month (measured value when Gemini 3.0 Flash was tested for one day per month over 30 days).

Optimized Engineer Assignment Based on Skills and Streamlined Inquiry Response Operations

Since the problematic code is visualized and identified in advance, engineers with appropriate experience can be assigned. Furthermore, for routine inquiries from the customer support team, the efficiency of problem identification, which previously took half a day, has been significantly improved by linking Claude, New Relic, and GitHub via New Relic MCP Server.

airCloset is currently working on implementing "Graph RAG," which enables AI to structurally understand source code. They are independently developing MCPs for "Code Graph," which analyzes source code dependencies, and "DB Graph," a dictionary of database table structures. In the future, they aim to complete the cycle of "problem detection → cause identification → fix code generation → code review → testing → deployment" as an automated pipeline using these MCP foundations. They plan to achieve autonomous operation by 2026, where AI can handle everything from detection to the release of corrected code for some performance issues, by preparing a state where AI can review code corrected by AI. They are also considering automation of security operations and the integration of all internal data sources with New Relic as a consolidated platform.

Comment from Ryosuke Tsuji, Executive Officer, CTO, and Product Group Manager at airCloset, Inc.

"Our company aims to maximize the 'value of time' for each customer and employee. In that sense, the data accumulated in New Relic is a treasure trove. If we can appropriately transfer this data to AI through our MCP server, we believe we can automate unproductive tasks such as routine triage and failure investigation, allowing engineers to focus on more creative development and thereby enhance the value of people's time. We intend to deepen our use of New Relic, combine the vast data accumulated in our observability platform with AI, promote autonomous operation of our core systems, and create an environment where people can focus on exciting and creative work."

For details on the "airCloset" New Relic adoption case, please see:

https://newrelic.com/jp/customers/aircloset

The URL for this press release is:

https://newrelic.com/jp/press-release/20260708

Other New Relic customer adoption cases can be viewed here:

https://newrelic.com/jp/customers

New Relic fact sheets, logos, etc., can be found here:

https://newrelic.com/jp/about/media-assets

About New Relic

Founded in 2008, New Relic is an industry leader providing an "observability platform" that makes all critical metrics of digital businesses observable. By observing not only the applications and infrastructure that make up digital businesses but also the customer experience from the user's perspective, New Relic enables companies to create DevOps teams that detect failures in digital services, detect degradation in customer experience, and identify and resolve potential problems and bottlenecks early. This allows companies to transform their digital transformation initiatives into measurable strategies. New Relic has over 16,000 customers worldwide, is used by more than half of Fortune 100 companies, and supports the digital transformation of hundreds of companies in Japan. Learn why New Relic is the preferred choice at newrelic.com/jp.

Features of the "New Relic" Observability Platform

As a leader in observability, New Relic supports engineers with a data-driven approach to planning, building, deploying, and running superior software. New Relic provides the only integrated data platform that allows engineers to acquire all telemetry (metrics, events, logs, traces) and, combined with powerful full-stack analysis tools, enables engineers to do their best work using data rather than opinions. New Relic offers a simple and transparent pricing structure. It provides high cost-effectiveness to engineers by supporting improvements in development cycle time, change failure rate, release frequency, and mean time to recovery (MTTR).

* New Relic is a registered trademark of New Relic, Inc.

* All product and company names in this document may be trademarks of their respective owners.