OpenAI cut monthly-close reporting from 5 days to 5 hours with Codex
OpenAI’s own finance team used Codex, the company’s AI coding agent, to compress one of their most painful monthly-close workflows from roughly five days to roughly five hours.
The project was led by Kyle Kober, OpenAI’s director of product finance, who focuses on consolidating product-level forecasts across user metrics, revenue, and compute demand. The specific bottleneck he tackled: reconciling product-usage data with accounting records for computing capacity costs, then turning that reconciliation into analysis and reporting.
How it happened
Kober said the finance team picked up Codex after it spread through OpenAI’s engineering organization around November or December of last year. Finance followed a couple of months later. Using Codex, the team automated enough of the reconciliation and reporting work to cut the process from about five days to roughly five hours.
The broader context is OpenAI CFO Sarah Friar’s push to build an “AI-native finance function,” which includes a goal of moving toward a zero-day close and continuously updated automated forecasting.
The wider trend
This fits a pattern BCG documented in a June 2026 report: AI coding agents can give finance teams a way to build targeted tools for analysis, matching, and anomaly detection without waiting in long development queues. Gartner predicted last year that 90% of enterprise software engineers will use AI code assistants by 2028, up from less than 14% in early 2024.
The risks are real too. Gartner has warned that AI coding costs could surpass the average developer’s salary by 2028 as token consumption rises and vendors shift toward consumption-based pricing. A March 2026 paper from the Cloud Security Alliance flagged that organizations are deploying AI-generated code into production systems despite documented security risks. BCG specifically called out auditability as an area where finance leaders need clear guardrails before going further.
The operator takeaway
If a finance team at OpenAI is manually reconciling usage data against accounting records and treating it as a multi-day project, smaller operators almost certainly have the same problem in some corner of their business. The pattern here is straightforward: find the workflow that takes days, involves data reconciliation, and produces a recurring report, then point an AI coding agent at it. The tooling is available now. The guardrails question is the part worth thinking through before shipping anything to production.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
