OpenAI employs over 100 former investment banking professionals to teach its artificial intelligence systems how to create financial models, aiming to eliminate the tedious work typically done by entry-level bankers throughout the sector.

This team, comprising past employees from major firms like JPMorgan Chase & Co., Morgan Stanley, and Goldman Sachs Group Inc., participates in a confidential initiative within the company called Mercury.

These contractors receive $150 hourly to develop prompts and construct financial models for various deal types, such as corporate restructurings and initial public offerings, based on information from someone knowledgeable about the program. OpenAI has also provided these workers with preliminary access to the AI system being developed to handle basic investment banking responsibilities.

Also Read: Chameleon Phones, Solar Charging Laptops, AI Ring - MWC 2025

This initiative highlights Sam Altman's OpenAI's pressing need to enhance its advanced AI capabilities for business applications across numerous sectors, spanning consulting, finance, legal services, and technology. Even after achieving a $500 billion valuation this month, the world's most valuable startup remains unprofitable.

Junior investment banking analysts commonly work over 80 hours weekly at their workstations during active transactions, creating comprehensive models using Microsoft Corp.'s Excel software for deals like mergers and leveraged buyouts. They frequently receive continuous demands from senior staff to modify PowerPoint presentations, followed by additional revisions — a practice that has generated Wall Street's famous "pls fix" internet meme.

Multiple startups are already positioning themselves to provide banks with AI solutions for these tasks. Though analysts have historically criticized the monotonous nature of their work, artificial intelligence advancement now raises questions about employment stability.

Also Read: India to Diversify the Semiconductor Supply Chain through Semicon 2.0

The initial stage consists of approximately a 20-minute conversation with an AI chatbot that poses questions related to the candidate's resume. The second round evaluates applicants' understanding of financial statements. The concluding phase involves a modeling examination.

The position offers flexibility, with contractors required to deliver one model weekly, according to reports. Guidelines involve crafting prompts using straightforward language, followed by model implementation. Contributors obtain feedback from evaluators and must address any problems before their output is integrated into OpenAI's infrastructure.

Reports indicate that Project Mercury has attracted individuals with prior experience at various Wall Street firms, such as Brookfield Corp., Mubadala Investment Co., Evercore Inc. and KKR & Co. Several current MBA students from Harvard University and Massachusetts Institute of Technology are also taking part in this initiative.

Also Read: The Quantum Computing Age is Upon Us Sooner than Expected

Contributors must develop their models using Excel while adhering to standard industry formatting practices, including specifications for margin dimensions and italicizing percentages.