Oracle reports faster recruiting and operations with ChatGPT and Codex
Oracle says more than 100,000 employees are using ChatGPT Work and Codex across talent acquisition, the Oracle Applications Lab and IT. The company reports a 98% reduction in the time needed for talent-acquisition research, 130,000 active ChatGPT users and more than 95,000 active Codex users.
In recruiting, Oracle’s talent acquisition team built a talent market intelligence tool with ChatGPT Work. It takes a job description, researches comparable roles, benchmarks compensation and assesses the talent pool in relevant locations. Jan Ackerman, Oracle’s Senior Vice President and Global Head of Talent Acquisition, said the information previously took two to four days to compile and can now be prepared in about 15 to 20 minutes.
Standardising recruitment intelligence
The speed gain is paired with a more consistent intake process. Oracle says recruiters had approached searches differently, whereas the new tool gives hiring managers a common quality of data and insights regardless of which recruiter supports the search.
The deployment also follows enterprise Codex adoption described by enterprise Codex adoption at scale as organizations seek to scale governed use of the coding system rather than leave teams to work in isolation.
Oracle’s figures are presented as active-user counts, spanning business and technical work rather than a single department. The company frames the change as moving specialist-dependent tasks into repeatable workflows that can be completed much faster.
From plain-language questions to operational work
Oracle Applications Lab built an ontology describing company objects, relationships and rules. With that structure in place, a business user can state the desired outcome in plain language, while Codex determines which internal systems to call, gathers the information and returns an analysis, report or application.
Richard Lam, Group Vice President at Oracle Applications Lab, described a user receiving an answer almost immediately to a question that would normally take several hours. When the result was checked against the previous manual process, he said the figures matched exactly.
In production engineering, site reliability engineers use Codex to assemble relevant incident context and bring up the appropriate playbook. Lam said a typical simple incident that previously took an hour to resolve can be handled in minutes, allowing SREs to spend less time locating information and more time guiding decisions.
Human responsibility remains central
Oracle does not describe these workflows as autonomous. Lam said teams must remain responsible for system design, architecture, security and the way Codex is asked to structure code. He also warned that teams need to work alongside Codex so that generated code remains maintainable.
Barry Shilmover, Vice President and Technical Advisor to the CIO, said prototypes increasingly replace written specifications when teams express an idea. For businesses assessing similar deployments, Oracle’s experience points to a practical requirement: define the workflow and guardrails, maintain the underlying systems, and keep people accountable for the output.