NTT DATA Group cuts incident analysis to 30 minutes with Codex

NTT DATA Group is using ChatGPT Enterprise and Codex so about 9,000 employees can automate day-to-day work. The reported result is incident analysis reduced to 30 minutes, with the same stack used to push secure AI adoption across the organization rather than limiting it to a small pilot group.

On the product side, the setup pairs ChatGPT Enterprise for general workplace automation with Codex for software-oriented tasks such as reading code, proposing fixes, and supporting investigation steps that usually sit in tickets and runbooks. That split lets non-coding staff automate routine analysis and documentation while engineering and operations teams use Codex closer to the systems under incident review.

For engineers and builders, a 30-minute incident analysis target changes how triage, log review, and root-cause write-ups are paced. If model-assisted investigation is trusted enough to sit in the critical path for a large services firm, teams need clear handoffs: what the model may draft, what a human must verify, and how findings enter the official incident record without inventing evidence.

The competitive angle is less about a single feature and more about scale and control. NTT DATA Group is framing ChatGPT Enterprise plus Codex as a way to reach thousands of employees under enterprise security constraints, which is the same adoption problem other large IT and consulting firms face when they try to move past shadow AI tools. Vendors that cannot show both measurable ops outcomes and a path to broad, governed rollout will struggle to match that pitch.

The practical takeaway is to treat incident analysis as a concrete pilot metric, not a vague productivity goal. Watch whether NTT DATA Group keeps the 30-minute analysis figure as usage grows past the initial 9,000-employee base, and whether Codex stays limited to coding workflows or becomes standard gear in production incident response.