FP&A / Finance Expert (AI Training & Evaluation)

A leading AI lab is expanding its roster of expert contributors, and Braintrust is helping identify experienced FP&A and operational finance professionals to support the development and evaluation of its advanced AI systems. This is a remote, flexible, freelance engagement where you apply your professional expertise to help make AI more accurate and reliable on complex finance topics.

In this role you will create technical, domain-specific prompts and review AI-generated responses end to end — the final output, the underlying reasoning, and the steps taken to get there — researching and verifying their accuracy against reliable sources and providing clear, expert-level feedback. Your judgment helps define what a correct, complete, and well-reasoned answer looks like in financial planning, analysis, and assurance. It is an opportunity to contribute to frontier AI work while leveraging deep domain experience, on a schedule that fits alongside your existing commitments.

Responsibilities:

• Create and refine technical, domain-specific prompts that probe and stress-test the AI's capabilities in FP&A / operational finance.
• Evaluate AI-generated responses end to end — the final output, the underlying reasoning, and the intermediate steps — for accuracy, completeness, and quality.
• Research and verify figures, methods, and claims against authoritative sources.
• Identify errors, gaps, edge cases, and inconsistencies in AI outputs, and clearly explain what is wrong and why.
• Apply detailed project guidelines consistently, and flag ambiguity or areas where guidance could improve.
• Provide expert written feedback that improves how AI systems reason about finance.
• Maintain a high standard of accuracy and consistency across all assigned work.

Requirements:

• 8+ years of professional experience in FP&A or operational finance & assurance (financial planning & analysis, budgeting/forecasting, controllership, or audit/assurance).
• Deep, demonstrable specialization in the above.
• Strong written English and the ability to explain complex reasoning clearly.
• Excellent attention to detail and sound analytical judgment.
• Comfortable working independently in a remote environment.
• Based in the United States.
• Availability for approximately 4 hours per day, ~5 days per week including weekends, for a short-term engagement.

Preferred:

• Prior experience reviewing, evaluating, or quality-checking content, data, or AI/model outputs.
• Familiarity with AI or machine learning projects.