NTT DATA Group cuts incident analysis to 30 minutes with Codex
NTT DATA Group has rolled out ChatGPT Enterprise and Codex companywide, with around 9,000 employees using AI to automate their work. In a flagship example, Codex cut incident analysis that previously took five engineers three days down to just 30 minutes.
Details
- Background: Following a May 2025 strategic partnership with OpenAI, the company is shifting its growth model from headcount expansion toward AI-driven value creation
- Deployment structure: NTT DATA established an internal “OpenAI Center of Excellence” to handle license distribution, technical validation, internal events, and knowledge sharing
- ChatGPT Enterprise results: Following companywide rollout, 96% employee satisfaction and 95% reporting productivity gains
- Codex scale: Used by roughly 9,000 employees across both technical and nontechnical roles, functioning as an AI agent that independently investigates, executes, tests, and revises based on instructions
- Key results: Incident analysis cut from three days of work by five engineers to 30 minutes; weekly active users grew 1.4x following training and documentation efforts
- Other use cases: Automated extraction of travel expenses from credit card statements, Excel data analysis by nontechnical staff, and analytical report generation without BI tools
- Governance: Guidelines cover data usage, system connectivity, network traffic, sandbox modes, automation levels, and human review requirements, giving the organization confidence in broad deployment
How to try it
This is a case study of NTT DATA Group’s internal deployment, not a consumer product announcement. Companies considering ChatGPT Enterprise or Codex for their own organization can refer to OpenAI’s official announcement or the Codex product page.