From days to a short recruiter briefing
Oracle says its talent-acquisition team built a market-intelligence tool in ChatGPT Work. Given a job description, it researches comparable roles, benchmarks compensation and assesses the talent pool across relevant locations. The company says that preparation previously took two to four days; recruiters now spend about 15 to 20 minutes using the tool before meeting a hiring manager. OpenAI lists this as a 98% reduction in research time.
The value is not simply faster text generation. Oracle describes a repeatable workflow that gives different recruiters a consistent starting point for the same kind of search. That can make intake conversations better prepared, while leaving the recruiter responsible for interpreting the market and deciding what to do next.
Codex moves beyond code generation
The same case study describes Codex translating plain-language questions into SQL against an internal business ontology. Oracle says users can request an outcome and receive an analysis, report or application from connected systems. One employee reportedly checked an AI-produced answer against the former manual process and found matching numbers. Oracle also says site-reliability engineers use Codex to gather incident context and surface playbooks.
These are company-reported examples, not an independent audit of every workflow. Oracle explicitly says the systems are not autonomous: people still need to design the architecture, apply security controls and maintain the code. That distinction matters. The practical change is that more employees can reach specialist workflows, while teams retain responsibility for whether the underlying data and rules are trustworthy.
What to watch in deployments
Oracle reports about 130,000 active ChatGPT users and more than 95,000 active Codex users. Those adoption figures show reach inside one large enterprise, but they do not tell us which workflows are most valuable or whether the time savings persist across all roles.
The revealing measure is whether a workflow can be repeated safely: how often users verify the answer, whether the data source is current, and how much of the old process remains. Oracle’s recruiting example makes the shift easy to picture: a two-to-four-day research packet becomes a short preparation session. The harder question is whether that speed also improves hiring decisions.
Source published 2026-10-08. Coverage is based on the maker’s announcement and demonstration.