Goldman Sachs is revolutionizing how its employees access decades of accumulated financial data through a new generative AI chat interface called Legend AI Query. The tool allows Goldman employees to ask questions in plain English and receive answers drawn from the bank’s vast data repositories, including information about client transactions, trades, investments, and loans, combined with external data from Bloomberg and Nasdaq.
According to Goldman’s Chief Data Officer Neema Raphael, the AI-powered search tool can surface information that users didn’t even know existed, creating “information superintelligence” that helps employees build better mental models faster. This represents a major shift for banks, where valuable data has historically been siloed and accessible only to those who knew exactly where to look.
Goldman isn’t alone in this AI-driven search revolution. JPMorgan’s private-bank AI copilot helps advisors track information in real time, Bank of America’s Banker Assist aggregates internal and third-party data for employee insights, and Morgan Stanley’s AIMS helps advisors search internal content. Blackstone recently spent 10 months building its own AI-powered search engine, with much of that time dedicated to solving permission and access control issues.
The challenge of enterprise search in finance is significant. Financial terminology is complex and nuanced — words like “hedge,” “ticker,” and “options” have different meanings outside financial contexts, making off-the-shelf products like ChatGPT less effective without customization. Balyasny Asset Management hired former Google AI scientist Peter Anderson to improve its systems, resulting in a 60% improvement in surfacing relevant documents after training OpenAI’s models on financial jargon.
Goldman has already launched two generative AI tools: a developer copilot that increased coding efficiency by roughly 20%, and the new Legend AI Query for data search. The bank also recently launched Legend Copilot to help data engineers manage and organize data more efficiently. According to Keri Smith from Accenture, who helps financial firms with AI strategy, effective search is just the foundation for more complex AI applications.
A new class of fintech startups is emerging to serve Wall Street’s AI needs. Rogo has onboarded about 25 Wall Street firms to its platform offering AI assistants capable of junior-banker-level duties. Meanwhile, Mako, founded by two Stanford graduates, recently raised $1.55 million from an early OpenAI backer to build a generative AI associate for private equity.
Morgan Stanley’s head of firmwide AI, Jeff McMillan, emphasized that enterprise search saves time for innovation while lowering barriers to accessing the firm’s intellectual capital, essentially “arming employees with knowledge firepower for meetings and discussions.”
Key Quotes
gives you this sort of information superintelligence to help the human build a better mental model faster and quicker with more sources
Neema Raphael, Goldman Sachs’ Chief Data Officer, explained how Legend AI Query transforms employee capabilities by surfacing previously inaccessible data, fundamentally changing how bankers can leverage the firm’s accumulated knowledge.
The power of enterprise search lies in its ability to save time so that humans can innovate and interact. Further, it lowers the barriers for employees to access robust intellectual capital quickly from the firm’s top experts, essentially arming them with knowledge firepower for meetings and discussions.
Jeff McMillan, Morgan Stanley’s head of firmwide AI, articulated the strategic vision behind Wall Street’s search initiatives, emphasizing that the goal extends beyond efficiency to fundamentally empowering employees with institutional knowledge.
Firms are realizing the value of enterprise search is not just a typical search engine but for all the downstream applications you’re going to be able to build on top of it
Gabe Stengel, CEO of fintech startup Rogo, highlighted that effective search is merely the foundation for more sophisticated AI applications, explaining why his company has successfully onboarded 25 Wall Street firms to its platform.
It’s actually just fundamentally a difficult problem to suck out and then rank what might be useful
Neema Raphael acknowledged the technical challenges of enterprise search, explaining why even sophisticated financial institutions have struggled with this problem despite Google solving consumer search decades ago. Personalization, permissions, and relevance ranking remain complex hurdles.
Our Take
Wall Street’s embrace of generative AI for search represents a watershed moment that validates the technology’s practical enterprise value beyond the hype. What’s particularly striking is how financial institutions are approaching this as infrastructure investment rather than experimental projects — Goldman’s 20% efficiency gains and Balyasny’s 60% improvement in document retrieval demonstrate measurable ROI.
The emergence of specialized fintechs like Rogo and Mako is equally significant, suggesting a new AI services layer is forming specifically for financial services. This mirrors the early cloud computing era when industry-specific solutions proved more valuable than generic platforms.
Most importantly, these search tools are Trojan horses for broader AI transformation. Once firms perfect data access and contextual understanding, they’ll unlock more sophisticated applications — automated analysis, predictive modeling, and eventually AI-driven decision-making. The banks investing heavily in search infrastructure today are positioning themselves to dominate the AI-powered finance landscape of tomorrow. The competitive advantage will be enormous.
Why This Matters
This development marks a pivotal moment in Wall Street’s digital transformation, as the world’s most powerful financial institutions finally crack the code on accessing their most valuable asset: data. For decades, banks have accumulated massive repositories of financial information but lacked effective tools to leverage it. Generative AI is changing that equation fundamentally.
The implications extend far beyond simple productivity gains. Perfecting enterprise search is the foundation for more sophisticated AI applications that could automate complex financial analysis, risk assessment, and investment decisions. As these tools mature, they’ll likely reshape the skills required in finance, potentially reducing demand for junior analysts while increasing the value of strategic thinking and client relationships.
For the broader AI industry, Wall Street’s adoption validates generative AI’s enterprise value and demonstrates that specialized, domain-specific AI applications often outperform general-purpose tools. The willingness of firms like Balyasny to hire Google AI scientists and the emergence of specialized fintechs like Rogo and Mako signal that financial services will be a major driver of AI innovation and investment. This could accelerate AI development across other data-intensive industries facing similar search and knowledge management challenges.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources: