Trade Validation Down from 30 Minutes to Under 4: How Chatham Financial Uses Codex and GPT-5.6

Hi everyone, it's Shii-chan! Today is an OpenAI customer story, and the numbers are eye-catching.

OpenAI News openai.com

What was announced?

OpenAI News shared a story about Chatham Financial, a firm that advises on capital markets decisions.

Chatham uses Codex to build internal and client-facing tools, and GPT-5.6 to power AI features. The work spans three areas:

• Employee-built applications (the internal platform Chatham Vibes)
• Chatham Onyx, its next-generation capital markets operating system
• Process Zero, its workflow reengineering consulting service

Why it matters

In finance, accuracy and auditability are paramount. Chatham wants to turn proven processes into governed, repeatable workflows and software, so experts spend less time on routine execution and more on client needs.

What changes

The headline example is trade validation:

• The Controls and Data Integrity team validates that each system record reflects what the client authorized and what was executed
• An app built with Codex gathers supporting transaction evidence, compares key terms, and flags discrepancies for review
• In early measurement, review dropped from approximately 30 minutes to under 4 minutes

The article stresses that Chatham is validating the app against real transactions and experienced reviewers before expanding automation.

Dive Deep

With Process Zero, for each workflow Chatham identifies the minimum inputs and evidence required, where human judgment is essential, and how AI and AI-built tools should support the work.

The model choices are spelled out:

• AI features in Chatham Vibes apps run on GPT-5.6 Terra by default, with GPT-5.6 Sol as an optional upgrade per app
• Chatham Onyx uses GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.4, and GPT-4.1
• Simple analysis and non-production testing go to the most cost-effective models, while GPT-5.6 handles complex tasks that need accuracy
• Codex is used across the Onyx lifecycle to plan, build, test, document, and review software

One Onyx example is ChatFIN, which summarizes patterns in historical market data, helps users understand their portfolios, and locates and links to legal documents for debt, derivative, and lease terms.

Next, Chatham plans to extend trade validation to more trade types and automate more of the workflow, with appropriate controls.

Wrap-up

• Chatham Financial builds internal and client tools with Codex and GPT-5.6
• Trade validation went from about 30 minutes to under 4 in early measurement (still being checked against real transactions and reviewers)
• Models like GPT-5.6 Terra and Sol are used according to the task

A good read if you want to see how AI fits into regulated, accuracy-critical work!