NTT DATA Group Cuts Incident Analysis to 30 Minutes with Codex

NTT DATA Group, a global IT services company based in Japan, has successfully slashed complex incident analysis time from three days to just 30 minutes using OpenAI's Codex. This achievement is not just a leap in technical efficiency, but also marks the expansion of AI's role from a conversational assistant to an autonomous agent capable of executing tasks independently in corporate environments.

From ChatGPT Enterprise to the Foundation of Codex

NTT DATA's journey began with the rollout of ChatGPT Enterprise across the company after a global strategic partnership with OpenAI in May 2025. They established an internal Center of Excellence (CoE) to manage license distribution, technical validation, use case development, and usage monitoring. As a result, over 96% of employees expressed satisfaction and more than 95% reported increased productivity in internal surveys.

The habit of using AI for research, writing, and content creation then became the foundation when the company introduced Codex to around 9,000 employees, both technical and non-technical.

Early Breakthrough: Incident Analysis from 3 Days to 30 Minutes

The first success story that caught leadership's attention was the automation of incident analysis on a critical system. Previously, this task required five experienced engineers and three full days. With Codex, the entire process was completed in 30 minutes.

"Codex has changed how people across the company view AI. The idea that AI can lead the execution of work has an impact similar to the emergence of ChatGPT," said Hiroaki Sato from the AI Technology Department of NTT DATA Group.

Codex's ability to not only write code but also investigate, execute, test, and revise based on instructions serves as tangible proof of the agentic value of AI in corporate settings.

Beyond Engineering: Non-Technical Employees Also Benefit

Following a 'Client Zero' approach that treats internal as the first customer, NTT DATA encouraged employees across various functions to test Codex. As a result, usage expanded beyond software development. Non-technical employees began using Codex to build lightweight tools, organize large volumes of files, analyze data in Excel, summarize documents, and automate repetitive processes.

A concrete example: an employee used Codex to extract transportation costs from credit card statements and transfer them to a travel reimbursement form. Codex worked across files, understood the structure of each sheet, and helped verify the transfer results, replacing a complex manual process. Now, employees without development skills can create analytical reports directly from raw data, reducing reliance on specialized business intelligence tools and dashboards.

"Codex is more than just helping developers write code faster. ... By enabling AI to research, organize, execute, and validate work, Codex empowers every employee to transform their approach to their roles," said Yuji Shono, Head of Global AI Office at NTT DATA Group.

Security and Governance Key to Large-Scale Adoption

To safely expand the use of Codex, the CoE published security and adoption practice guidelines. These guidelines clarify what data can be used, which systems can be connected, network traffic management, appropriate sandbox modes, the right level of automation, and when human review is necessary. This secure and managed environment allowed both technical and non-technical employees to explore Codex without worry.

"Creating a secure and well-managed environment is essential for employees to use Codex with confidence. Our vision is to make Codex more than just a tool for engineers. We want it to become a natural part of every employee's daily work," added Yuji Shono.

Measurable Results and Leadership Lessons

Some measurable achievements from Codex adoption at NTT DATA Group:

• Complex incident analysis completed in 30 minutes, down from 3 days by 5 senior engineers.
• Weekly active Codex users increased 1.4 times after the release of guidelines and hands-on training.
• Internal system operations automated using Playwright, cutting daily routine task time.

From this experience, NTT DATA distilled five key lessons for organizations looking to adopt Codex:

• Make ChatGPT Enterprise part of daily work to build the habit of collaborating with AI.
• Deploy ChatGPT Enterprise widely to create network effects through peer learning and word-of-mouth.
• Create an environment that allows employees to use AI safely with adequate security and privacy protections.
• Treat deployment as the beginning, then continuously improve the adoption program using usage data, surveys, and employee interviews.
• Have the CoE identify and generalize high-impact use cases, then share them as reusable best practices with appropriate safeguards.

Relevance for AI Users and Developers in Indonesia

NTT DATA's approach shows that successful AI adoption in companies requires more than just technology; it also needs a foundation of culture, security, and identification of high-impact use cases. For companies and developers in Indonesia, these lessons are relevant: starting by encouraging the widespread and safe use of AI tools can pave the way for more advanced task automation. Codex proves that autonomous AI is not only for engineers but can also empower non-technical roles, lowering barriers to completing work that previously required specialized expertise. This opens opportunities for local developers to integrate similar capabilities into workflows across various industries.