Updated
Updated · OpenAI · Oct 8
Oracle Deploys ChatGPT to 100,000 Workers, Cutting 2-4 Days of Tasks to Minutes
Updated
Updated · OpenAI · Oct 8

Oracle Deploys ChatGPT to 100,000 Workers, Cutting 2-4 Days of Tasks to Minutes

3 articles · Updated · OpenAI · Oct 8

Summary

  • More than 100,000 Oracle employees are using ChatGPT Work and Codex, with recruiting, analytics and engineering tasks that once took days or hours now completed in minutes.
  • Oracle’s talent acquisition team built a market-intelligence tool that turns a job description into pay, role and location benchmarks, shrinking recruiter prep from 2-4 days to about 15-20 minutes.
  • Oracle Applications Lab paired Codex with an internal ontology so business users can ask plain-language questions and get SQL-driven reports or analyses almost immediately; one user said the output matched the old manual process exactly.
  • Production engineers also use Codex during incidents, where simple outages that once took about an hour to resolve can now be handled in minutes by automatically surfacing context and the right playbook.
  • Oracle said the shift still requires human oversight on system design, security and code maintainability, even as AI becomes a default tool across the company’s workflows.

Insights

As AI accelerates coding, how will enterprise giants overcome the hidden new bottlenecks in testing and deployment?
If AI slashes task times from days to minutes, can human oversight truly keep pace without becoming the ultimate bottleneck?
When employees transition to mere prompters, what happens to the deep institutional knowledge previously built through manual problem-solving?