OpenAI's Project Mercury: Revolutionizing Investment Banking with AI

OpenAI looks to replace the drudgery of junior bankers' workload | Fortune

The group, which includes former employees of JPMorgan Chase & Co., Morgan Stanley, and Goldman Sachs Group Inc., is part of a secretive project inside the startup that's code named Mercury, according to documents seen by Bloomberg. Participants are paid $150 per hour to write prompts and build financial models for a range of transaction types, including restructurings and initial public offerings, according to a person familiar with the effort. The company has also granted the contractors early access to the AI it's creating that aims to replace entry-level tasks at investment banks. The project underscores the urgency at Sam Altman's OpenAI to make its powerful AI technology more useful to businesses across a wide swath of industries, from consulting to finance to legal to technology. Despite reaching a $500 billion valuation earlier this month, the world's largest startup has yet to turn a profit. A spokesperson for OpenAI said the company works with a range of experts "to improve and evaluate the capability of our models across different domains. Experts are recruited, managed and compensated by third party suppliers." Investment banking analysts typically spend upwards of 80 hours a week at their desks when working on live deals, building detailed models in Microsoft Corp.'s Excel program for mergers and leveraged buyouts alike. They often face a steady stream of requests from higher-ups to make tweaks to PowerPoint slide decks, and then tweaks to those tweaks -- a culture that's spawned Wall Street's "pls fix" meme. Already, a bevy of startups are looking to step in and equip banks with AI that can help with all that. While analysts have long complained about the drudgery, the rise of AI is now sparking concerns about their job security. Read more: Junior Bankers Say Grunt Work Matters as AI Takes On Hated Tasks The application process for Project Mercury involves almost no human interaction, according to the person familiar with the matter, who asked not to be named discussing non-public information. The first step is a roughly 20 minute interview with an AI chatbot, which asks questions based on the applicant's resume. The second phase tests candidates on their knowledge of financial statements. The final stage is a modeling test. The job is flexible and contractors are expected to submit one model per week, the person said. Instructions include writing prompts in simple terms, then executing the model. Participants receive feedback from a reviewer and are expected to fix any issues before their work is ultimately plugged into OpenAI's systems, the person said. Project Mercury has so far drawn participants who've previously worked at a variety of Wall Street outposts, including Brookfield Corp., Mubadala Investment Co., Evercore Inc. and KKR & Co., the documents show. Some current MBA candidates at Harvard University and Massachusetts Institute of Technology are also participating in the effort. Participants are asked to create their models in Excel and they're also expected to follow industry norms for formatting the models, including for areas like margin sizes and italicizing percentages.

OpenAI Is Paying Ex-Investment Bankers $150 an Hour to Train Its AI

The initiative arrives as OpenAI seeks to find real-world applications for AI technology. OpenAI has assembled a team of more than 100 former investment bankers to train its AI to automate the tedious, manual tasks typically performed by junior bankers. The group, which includes former employees of Morgan Stanley, JPMorgan Chase, and Goldman Sachs, is tasked with creating financial models for various transactions, including restructurings and initial public offerings, according to internal documents seen by Bloomberg. The initiative, code-named Project Mercury within OpenAI, pays contractors $150 per hour to feed financial models into AI, with the aim of expanding the real-world, practical use of AI across business sectors, like finance and technology. Bloomberg points out that despite reaching a $500 billion valuation earlier this month, cementing its status as the most valuable private company in the world, OpenAI has yet to turn a profit. Related: OpenAI Made More Money in the First Six Months of the Year Than It Did in All of 2024 Applying to Project Mercury is largely automated, beginning with a 20-minute interview with an AI chatbot, and progressing to tests on financial statements and modeling, sources told Bloomberg. Contractors who make it past the screening are required to submit one financial model per week in Excel, following industry standards. There was no public application open on OpenAI's careers site for Project Mercury at the time of writing. Project Mercury has reportedly attracted interest from participants who have previously worked on Wall Street, as well as current MBA candidates at Harvard University and the Massachusetts Institute of Technology, per Bloomberg. With Project Mercury, OpenAI aims to transform traditional investment banking workflows that depend on repetitive tasks. For example, junior investment bankers log 100-hour workweeks as they build detailed Excel models for mergers and make modifications to PowerPoint slide decks -- tasks that AI can help with. Related: Bank of America Is Cracking Down on Overwork for Junior Bankers and Capping Hours to 'Only' 80 a Week Still, the automation of grunt work presents both opportunities and concerns for junior bankers. Though automating modeling and presentation tasks may alleviate burnout, industry expert Jeanne Branthover, head of global financial services at recruiting firm DHR Global, told Bloomberg in March that performing these tasks manually helps junior workers develop industry knowledge, attention to detail and confidence for client interactions. "Reading the documents, analyzing them, there's a process that you need to learn," Branthover told the outlet. Branthover added that missing out on this kind of work is "going to be detrimental to the young bankers." Major financial and consulting firms are already adopting generative AI. Citigroup, for instance, began deploying its AI tool Stylus in late 2024 to 140,000 employees across eight countries. The platform can summarize, compare and search through multiple documents at once, drastically reducing the time spent on manual analysis. Meanwhile, McKinsey reported earlier this year that more than 75% of its 43,000 employees regularly use Lilli, its in-house AI platform introduced in 2023, to generate PowerPoint slides, research materials and draft client proposals.

OpenAI Is Coming After Entry-Level Finance Work

The ChatGPT maker's secretive "Project Mercury" aims to teach A.I. how to handle the tedious tasks once done by Wall Street juniors. Revising pitchbooks, formatting slides and tweaking font sizes -- these are some of the tedious tasks that have long fallen to entry-level investment bankers eager to climb the ladder at Wall Street's top firms. Now, former staffers across major banks are helping OpenAI train its A.I. to master such grunt work in a bid to expand the ChatGPT-maker's reach in the financial sector, according to Bloomberg. Sign Up For Our Daily Newsletter Sign Up Thank you for signing up! By clicking submit, you agree to our <a href="http://observermedia.com/terms">terms of service</a> and acknowledge we may use your information to send you emails, product samples, and promotions on this website and other properties. You can opt out anytime. See all of our newsletters OpenAI has reportedly hired more than 100 contractors who previously worked at firms like JPMorgan Chase, Morgan Stanley and Goldman Sachs to take part in a secretive effort known internally as Project Mercury. Participants are expected to submit one financial model per week and are paid around $150 per hour, according to Bloomberg, citing documents and sources familiar with the matter. The revelation is likely to heighten anxiety among Wall Street's junior bankers, as A.I. continues to take on work that was once the domain of entry-level employees. Earlier this year, Stanford researchers found that workers aged 22 to 25 experienced a 13 percent decline in employment across job sectors exposed to A.I. since late 2022. Like many other industries, Wall Street is rapidly embracing A.I., deploying it internally at firms from Morgan Stanley to Citigroup to Bank of America. Goldman Sachs, meanwhile, is piloting an A.I. software engineer nicknamed Devin, alongside other initiatives, including the expansion of its GS AI Assistant platform. JPMorgan, for its part, is investing $2 billion annually in A.I. Earlier this month, JPMorgan confirmed that its embrace of A.I. will lead to a slowdown in hiring. Due to A.I.'s "productivity tailwinds," the bank will "constrain people's headcount growth," Jeremy Barnum, JPMorgan's chief financial officer, told analysts during the bank's third-quarter earnings call. Goldman Sachs has expressed similar intentions. In a recent memo to staff, CEO David Solomon and other executives said the bank plans to "constrain headcount growth through the end of the year" and carry out a "limited reduction in roles" as it focuses on opportunities created by A.I. To apply for Project Mercury, prospective hires must first complete a 20-minute interview with an AI chatbot, according to Bloomberg. They are then tested on their financial statement knowledge before finishing the process with a modeling test. OpenAI regularly works with various experts "to improve and evaluate the capability of our models across different domains," the company said in a statement to the outlet, noting that such experts are "recruited, managed and compensated by third-party suppliers." OpenAI isn't the only A.I. developer aiming to boost its financial expertise. In July, Anthropic rolled out a suite of Claude tools tailored for financial services firms. Cohere, a Canadian A.I. rival specializing in enterprise applications, counts the Royal Bank of Canada among its clients. Elon Musk's xAI, too, is working to deepen its models' financial knowledge. Earlier this year, the startup said it would focus on hiring specialized domain experts to train its LLMs in fields like finance.

OpenAI hires Wall Street bankers to train AI in financial models

This content has been selected, created and edited by the Finextra editorial team based upon its relevance and interest to our community. The firm behind ChatGPT has been recruiting ex-bankers at the likes of JPMorgan and Goldman Sachs for its Mercury project, says Bloomberg, citing documents. The bankers are paid $150 an hour to write prompts and build financial models for things like restructurings and initial public offerings. To get the gig, they face an interview with an AI chatbot, followed by a test on financial statements, and a modeling test. Once recruited as contractors, participants must create models in Excel and follow industry norms for formatting. OpenAI is one of many firms looking to help Wall Street reduce the heavy manual workload required of investment banking analysts who build models for deals.

OpenAI Turns to Wall Street as AI Training Help | PYMNTS.com

By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions. According to Bloomberg, the company has hired more than 100 former investment bankers from firms such as Goldman Sachs, JPMorgan and Morgan Stanley to help its systems learn how to build and interpret financial models. The initiative, called Project Mercury, pays contractors about $150 an hour to create Excel models for initial public offerings, restructurings and leveraged buyouts, and to refine the model's outputs based on their expertise. The project marks a shift for OpenAI Chief Executive Sam Altman, who has been trying to move the company beyond consumer chatbots and into enterprise use cases that generate steady, high-value revenue. OpenAI's systems are already used for drafting documents and code, but those applications rely on general-purpose data drawn from the internet. Training on real-world financial work instead of open web text gives the company access to a structured and auditable data source that could make its AI models more accurate and commercially viable for corporate customers. Investment banking offers a type of data that large language models have struggled to learn from elsewhere. Analysts at major banks often spend upward of 80 hours a week producing spreadsheets that follow strict accounting conventions and logic chains. Training a model to understand those relationships is far more difficult than teaching it to summarize text. Public data doesn't contain the formulas or dependencies that define financial models, and genuine deal models are rarely available outside of firms' internal systems. Recent PYMNTS coverage on research by NYU Stern and FinTech firm Goodfin illustrates that limitation. The study found that while advanced AI models can now pass mock versions of the CFA Level III exam, one of the most demanding tests of analytical and ethical reasoning in finance, researchers warn that LLMs still do not think like analysts. OpenAI's decision to recruit former bankers is an attempt to generate high-quality training data that captures how professionals think about financial cause and effect. A model that learns how debt, cash flow and equity interact can be fine-tuned for valuation, risk or performance analysis later on. Even advanced language models struggle with multi-step calculations and logic consistency. Spreadsheet reasoning is hierarchical and sensitive to small errors. Teaching an AI system to keep those relationships balanced requires access to verified, expert-built examples. That is what Project Mercury is designed to provide: thousands of realistic, audited models built to professional standards. The move also reflects a broader challenge across the artificial intelligence industry. Models are nearing the limits of what they can learn from open data, creating pressure to source proprietary, high-fidelity training material. Finance, consulting and legal services produce datasets that combine quantitative structure with professional reasoning. For OpenAI, that data could underpin more reliable enterprise products suited for regulated environments. OpenAI is not alone in the shift toward domain-trained systems. Scale AI has restructured its data-labeling business to emphasize expert-level annotation in fields such as medicine, robotics, and finance. The company's blog describes a move away from large-volume, general labeling toward curated, expert-driven data pipelines designed for foundation-model training. That mirrors the strategy behind Project Mercury, where OpenAI uses domain specialists to create high-signal examples instead of synthetic or crowdsourced data. Snowflake has also entered the domain-specific LLM space with its Arctic model, an open-source system designed for enterprise workloads such as SQL generation, coding and data analysis. Arctic integrates directly into the company's data-cloud platform and is trained for structured reasoning within enterprise contexts rather than general conversation. Snowflake's goal is to build an AI layer optimized for accuracy, compliance, and enterprise reliability, the same attributes OpenAI seeks to achieve through targeted domain training. Other developers are following the same approach. Voyage AI has released finance-specific embedding models that outperform general embeddings on banking data. Research benchmarks show that even the most capable general-purpose models struggle on specialized reasoning without domain-specific fine-tuning.