Cisco & OpenAI: How Codex is remaking enterprise engineering
Curated by the Inblix editorial team
Cisco isn’t just using Codex as a fancy autocomplete tool—they’ve fully baked OpenAI’s AI into the core of how they build software. The results are eye-popping: over 95% of new AI features are now written by Codex, defect resolution is up 10-15x, and they’re saving 1,500+ engineering hours monthly. For AI Defense, Cisco’s big AI security product, Codex compressed work that usually takes multiple quarters down to just weeks. The real shift here? Codex acts like a full-fledged engineering teammate that can understand massive codebases, run autonomous compile-test-fix loops, and play nice with enterprise security and governance. Cisco even helped OpenAI tune Codex for the real world. Why it matters: This isn’t a toy demo—it’s a concrete blueprint showing that AI-native development isn’t just for startups, and it’s probably coming to an enterprise near you soon.
💡 Key Takeaways
- Cisco integrated Codex into production engineering workflows, not just as a developer productivity booster, achieving over 95% of new AI features written by the AI.
- Codex compressed feature delivery for Cisco's AI Defense product from quarters to weeks, saving over 1,500 engineering hours per month.
- The true value came from Codex's agency—its ability to reason across complex repos, run autonomous fix loops, and fit into enterprise security frameworks.
Keep reading: See related articles below for more coverage on this topic.
Get smarter about AI
The sharpest AI news, curated daily. Delivered free to your inbox.