OpenAI finance exec builds custom tools with Codex
OpenAI finance executive Kyle Kober used the company's own Codex AI coding agent to automate a monthly-close reporting process, cutting the time required from roughly five days to about five hours. The project shows how finance teams can now build targeted internal tools without waiting for IT or engineering support.
"There's now the opportunity to just do it yourself," Kober told CFO Dive in an interview.
Kober, who serves as director of product finance at OpenAI, focused the effort on streamlining reporting and analysis around the company's computing capacity costs. The work involved reconciling product-usage data with accounting records and converting that into analysis and reporting - a labor-intensive process that had long been one of the most cumbersome parts of monthly close.
How the automation took shape
Kober said the finance team began using Codex after the tool caught on among OpenAI's engineers. "I think most engineers had their Codex moment in like November or December of last year; it was only a couple of months later for us," he said.
Before joining OpenAI, Kober led product and corporate finance at Nextdoor. He previously worked as an associate at Financial Technology Partners, an investment banking firm focused on fintech.
The Codex project is part of a broader push across OpenAI's finance organization. In a blog post last month, OpenAI CFO Sarah Friar outlined plans to build an "AI-native finance function," including a move toward a "zero-day close" and automated, continuously updated forecasting.
AI coding gains ground in finance
The shift at OpenAI reflects a wider trend. 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. A June report from Boston Consulting Group said AI coding agents can give finance teams a way to build targeted applications for analysis, matching, and anomaly detection "without waiting for long development queues."
For finance professionals looking to build these skills, AI coding courses cover the fundamentals of working with tools like Codex. Similarly, AI for finance training focuses on applying automation to reporting, forecasting, and close processes.
The productivity gains come with new costs and risks. Gartner has warned that AI coding expenses are rising as token consumption increases and vendors shift toward consumption-based pricing. The firm predicts AI coding costs could surpass the average developer's salary by 2028. Tokens are the basic units of data processed by AI, and greater token consumption generally means higher costs.
A March paper from the Cloud Security Alliance said organizations are integrating AI-generated code into production systems at scale despite documented security risks. BCG cautioned that finance leaders need clear guardrails in areas such as auditability when bringing AI coding into their teams.
Why this matters for finance professionals
The OpenAI example sets a concrete benchmark: a five-day monthly process reduced to five hours using an AI coding agent. Finance teams that learn to build their own automation tools can cut close times and free capacity for higher-value analysis, but they need to manage the associated costs, security risks, and auditability requirements before scaling these tools across their organizations.
Your membership also unlocks:
