Goldman Sachs’ Journey in Implementing Generative AI Solutions


Artificial intelligence (AI) developments are evolving at a pace that is exceeding corporate technology leaders’ expectations. One significant area of focus is large language models (LLMs) and generative AI. These advancements hold the potential to automate tasks such as document categorization, coding, and information summarization. In turn, this can lead to increased productivity and efficiency.

Goldman Sachs Group Inc. Chief Information Officer, Marco Argenti, described the rapid pace of AI evolution as something that challenges their ability to comprehend and integrate it effectively. He suggests that the limitations are not in the technology but in humans’ capacity to rationalize and utilize it.

According to a KPMG LLP survey, 65% of executives believe that generative AI will significantly impact their organizations in the next three to five years. This belief extends to a broader societal impact, with 77% of respondents suggesting that AI’s influence would exceed other emerging technologies. However, adoption is lagging due to obstacles such as talent acquisition, cost, and data privacy concerns, with 60% of respondents suggesting they are a year or two away from implementing generative AI solutions.

Harnessing AI for Increased Efficiency at Goldman Sachs

Check out our weekly crypto and fintech newsletter here! Follow CryptoMode on Twitter, Youtube and TikTok for news updates!

Argenti, who transitioned to Goldman Sachs from Amazon Web Services in 2019, has a more optimistic timeline for implementing generative AI solutions. Goldman has several generative AI proof of concepts in place, with Argenti suggesting that full implementation could be closer to months rather than years.

Under Argenti’s direction, Goldman Sachs is using AI technology to automate coding, which he believes will not only result in efficiency gains but also free developers to concentrate on the aspects of their work that are more crucial for their clients.

Further, the financial giant is exploring other AI applications such as document classification and categorization, as well as information summarization. Argenti indicated that these experiments have shown promising results, including coding efficiencies in the double digits and document classification accuracies comparable to those achieved by humans.

Goldman Sachs’ Venture into AI-Powered Networking

In another innovative application of AI, Goldman Sachs has spun out the first startup from its internal incubator. This startup, Louisa, is an AI-powered networking platform intended for corporate use. Rohan Doctor, founder and CEO of Louisa, describes the platform as a sophisticated LinkedIn driven by AI. Louisa automatically generates user profiles and connects people who might benefit from knowing each other based on the data it gathers.

Louisa, part of the inaugural class of Goldman’s incubator program, was conceived to address the challenges of establishing connections within large corporations. Doctor believes that the lack of efficient networking within professional services firms like Goldman Sachs results in missed opportunities and fractured client experiences.

This AI-powered platform is especially timely given the recent surge in remote and hybrid work arrangements. Doctor posits that platforms like Louisa are necessary in a world where traditional networking methods are becoming obsolete.

Goldman Sachs’ exploration of generative AI and its spin-off AI-powered social media startup illustrates the company’s commitment to leveraging AI technology. These initiatives have the potential to redefine business operations and productivity, opening up new possibilities for increased efficiency, accuracy, and networking.

None of the information on this website is investment or financial advice. CryptoMode is not responsible for any financial losses sustained by acting on information provided on this website.