GenAI to become immersive across our businesses: Goldman Sachs' Samtani

Samtani talks about Indian engineers and Generative AI (GenAI) in banking

US investment bank Goldman Sachs was an early mover in India's technology ecosystem, setting up an office in Bengaluru in 2004 for information technology and back-office support. In two decades, its operations have grown to cover a wide range of banking functions. GUNJAN SAMTANI, co-chairman of Goldman Sachs, India, and country head of Goldman Sachs Services India, spoke with Avik Das in Bengaluru about Indian engineers and Generative AI (GenAI) in banking. Edited excerpts:

 

Goldman Sachs' technology division completed two decades in Bengaluru last year. How will you describe the journey?

 

At that time, we had a modest ambition that there would be, at some point, a few hundred people doing largely application support, infrastructure support, and some business execution for our operations and finance functions. We made some specific pivots along the way: Going from supporting business functions and technology applications and infrastructure to driving scale. Once we built the scale, we also developed expertise in products and businesses. Then came the pivot to develop deep, quantitative capabilities, which is a function called strats or quantitative finance.

 

And once we brought in strats, we started the journey of business process re-engineering. The transformation started as we started to build strats functions across each of our teams. The evolution led us to truly becoming front to back across each of our businesses – for entities, for almost all products of Goldman Sachs. That means it is no longer just about the traditional execution of business processes in the context of technology, finance, operations and strats. We have equally classic roles in wealth and asset management, sales and trading, and quantitative investment strategies.

 

What products or platforms developed in India have had an impact on Goldman's global ecosystem?

 

I will give an example but I wouldn't say rolled out on a standalone basis because we are a global organisation. One of the most sophisticated platforms for data analytics tied to pricing and risk management for commodities, such as oil and gas, metals and crude, was in a big part built out of our offices. The pricing and risk management of that data science is used by almost all our commodity traders and risk managers in the firm. There are many examples, including the work we did to build a platform for equity, quant trading platforms. We are talking about enabling platforms that operate at a latency of one millionth of a second.

 

How are you leveraging Gen AI?

 

There are three vectors in which we are looking at GenAI opportunities. The first is business growth and client service; the second is developer efficiencies; and the third is providing operating efficiencies at scale to the firm.

 

Because a third of our engineers are in India, the second bucket of developer efficiency truly matters to us. We have enabled a significant proportion of our developers on AI Copilot for software development. That's a journey that we took on. We recently rolled out GS AI Assistant to over 10,000 of Goldman Sachs employees globally. It is built on an AI platform that we designed and architectured at the firm. While not the entire team, many of the engineers who built both GS AI platform and GS AI Assistant are based in India. The purpose of GS AI platform is to take away the complexity of integrating with multiple LLMs (large language model).

 

What are the areas in your firm’s investment banking where you think Gen AI can be a path breaker?

 

I do anticipate AI to become immersive across many aspects of our businesses. Not limited to one or two but right from how we think about our banking workflows on M&A (mergers and acquisitions) and financing, but to how we think about wealth management and engagement with our clients as well as suitability analysis, to our businesses and asset management, and to how we think about risk compliance.

 

With GenAI playing a critical role across all levels, will global capability centres (GCCs) take a fresh look at hiring in the next 12 to 24 months?

 

I can only comment about our GCC. That said there is a macro report that GCCs will continue to hire based on the national policy that the Finance Minister has rolled out. The report also talks about the contribution of the GCC's increasing to over $105 billion in the next five years from $65 billion and headcount growing up to north of 2.5 million people from 1.9 million currently.

 

We have almost 1,000 open roles. About 40 percent of those are in engineering. We will see the pace of hiring as to the progression that we can make. Right now, we will look to deepen the expertise and the scale for the functions that we have built out of our offices in Bangalore (Bengaluru) and in Hyderabad.

 

What is the work at your two technology centres in Bengaluru and Hyderabad?

 

With the migration of many of the application and infrastructure support that started in India, along with the business execution, came the opportunity to transform the businesses on newer platforms for us to have a view into the bespoke flows that existed across our businesses in different regions, be it Asia or Europe or the Americas. That gave us an opportunity to horizontalise the business processes and standardize them on to common platforms for the firm. Of course, we worked with our global counterparts hand in hand for that business process re-engineering. In the process, we brought approaches of statistical finance for machine learning and analytical AI. That is how we were thinking about the transformation of our businesses globally – so we became a strong hub for providing that capability to the firm at scale.

 

Today, we are the second largest office of the firm with about 9,000 employees working in our offices in Bangalore and Hyderabad together. About half of those are engineers, which translates into a third of our global engineering talent.