Chatham Financial has overhauled its trade validation workflow using OpenAI models and coding assistants, cutting transaction review times from 30 minutes to under four minutes. The North American capital markets advisory firm deployed Codex and GPT-5.6 to automate repetitive compliance checks and build internal software, aiming to free its specialists to focus on high-judgment advisory work.
The time reduction stems from an application built for Chatham’s Controls and Data Integrity team, which verifies that executed transactions match client authorizations. Developed with Codex through the firm’s Process Zero workflow reengineering service, the tool collects supporting transaction records, compares contract terms, and highlights discrepancies for human oversight. According to figures published by OpenAI, early testing showed the system reduced review duration to less than four minutes.
Chatham also integrated OpenAI models into internal operations through an in-house development platform called Chatham Vibes. The system allows staff to build tailored utilities for daily research, drafting, and analysis tasks. Applications built inside Vibes run on GPT-5.6 Terra by default, with GPT-5.6 Sol available as a per-app upgrade. Employees have used the system to assemble fixed-income rate sheets, construct hedging dashboards, audit trade confirmations, and prepare pricing workbooks for maturing-cap trades.
For its flagship capital markets operating system, Chatham Onyx, the firm applied Codex across planning, testing, documentation, and code reviews. Onyx consolidates debt, derivatives, and asset records while maintaining source traceability. The platform routes simpler analysis and non-production testing to GPT-4.1 and GPT-5.4, while directing high-value tasks to GPT-5.6 Sol and GPT-5.6 Terra. One Onyx feature, ChatFIN, summarizes historical market patterns, analyzes client portfolios, and indexes legal contracts for debt and lease terms.
Firm executives framed the automation as an effort to reclaim analyst capacity rather than replace human oversight. Chief Executive Officer Matt Henry said the company designs workflows by identifying essential inputs, establishing where human judgment matters, and selecting supporting tools. Chief Technology Officer John DeGuenther added that Codex shortens product development timelines while maintaining existing security, accuracy, and accountability standards, and Co-Chief Operating Officer Mike Noonan noted that eliminating manual assembly gives advisors more time to handle complex client scenarios.
The expanded automation remains subject to ongoing verification. Alex Nordlinger, co-head of Chatham’s AI Advisory practice, said the firm is still comparing the trade validation application’s results against real transaction data and experienced reviewers before broadening deployment. Chatham plans to expand the validation software to additional trade varieties and continue developing new features for the Onyx platform.
