Oracle is accelerating internal workflows by deploying ChatGPT and Codex across recruiting, applications and IT teams, converting tasks that once required specialist effort and days of manual work into functions employees can complete in minutes. The company says the tools are already in use across talent acquisition, the Oracle Applications Lab and IT, with adoption reported among more than 100,000 employees.
In talent acquisition, Oracle built a market intelligence tool using ChatGPT Work that ingests a job description and automates research into comparable roles, compensation benchmarks and the talent pool across locations. Jan Ackerman, Senior Vice President and Global Head of Talent Acquisition at Oracle, says recruiters who previously spent two to four days compiling this information now prepare for hiring-manager conversations in roughly 15–20 minutes. Beyond time savings, the tool standardizes the intake process so every hiring manager receives consistent data and insights regardless of who handles a search.
Within the Oracle Applications Lab, engineers have used Codex to translate natural-language questions from business users into reliable SQL queries. The team created an ontology defining the company’s objects, relationships and rules so that a user can describe the desired outcome in plain language; Codex determines which internal systems to query, aggregates the data and returns an analysis, report or application. Richard Lam, Group Vice President at Oracle Applications Lab, said the change delivers near-instant answers for queries that previously took hours, and users have confirmed that the automated outputs match numbers produced by the prior manual process.
Site reliability engineers (SREs) have also adopted Codex to accelerate incident response. The tool gathers context about incidents and surfaces the correct playbooks, reducing the time SREs spend hunting for information and allowing them to focus on decision-making. Lam noted that a “typical simple incident that used to take an hour to resolve can now be handled in minutes.” This shift highlights how Codex is being applied to both analytics and operational functions.
Oracle’s rollout emphasizes that human responsibility and technical governance remain central as AI handles more operational tasks. Leadership across teams outlined three key lessons from their deployments: provide the correct guardrails with careful system design, architecture and security when structuring generated code; communicate ideas with working prototypes instead of paper specifications; and retain ownership of the code so generated artifacts remain maintainable over time. Barry Shilmover, Vice President and Technical Advisor to the CIO, said he now presents ideas as prototypes, and Lam warned that failure to maintain ownership of code produced by Codex could lead to unmaintainable artifacts.
The company reports thousands of ChatGPT and Codex users across multiple lines of business as it shifts operational practices to leverage these foundation models. Teams in recruiting, analytics and engineering are increasingly describing outcomes while letting Codex and ChatGPT handle implementation details, enabling individual contributors and small teams to deliver work that previously required larger specialist teams and longer timelines. That pattern has produced both speed gains and more consistent outputs across the organization.
Oracle’s experience also underscores a practical integration strategy for enterprises adopting foundation models: pair human oversight and governance with AI-generated output, build prototypes that demonstrate value quickly, and preserve developer ownership to prevent technical debt. Shilmover summarized the day-to-day impact by saying he does not know what his next problem will be, but he expects one of the first steps will be to leverage Codex in solving it.
Taken together, Oracle’s internal deployments of ChatGPT and Codex illustrate how large enterprises are embedding foundation models into business-facing and engineering workflows to accelerate work, improve consistency and preserve human accountability. The company’s reported reductions in research and incident-handling time, along with standardized processes and governance lessons, provide a practical example for other organizations evaluating the operational use of similar AI tools.
As Oracle continues to scale these models across lines of business, the firm’s account highlights both the operational benefits and the governance practices leaders say are necessary to sustain those gains without sacrificing maintainability or security.
Source: Read the original source
Leave a Reply