in retrospect In Nima Raphael’s 20-year career, there has probably never been a more important period than the present.
As the person responsible for processing vast amounts of data at Goldman Sachs, Raphael’s behind-the-scenes world of digital information has been brought to the forefront thanks to the excitement around AI.
“There is no AI strategy without a data strategy. We use this code all the time: garbage in, garbage out,” Rafael, chief data officer and Goldman partner, told Business Insider. “If you don’t understand that and it’s everywhere, your AI strategy won’t go anywhere.”
For banks like Goldman Sachs, data is virtually essential for everything from simplifying complex transactions for buy-side clients to helping investment bankers identify mergers and acquisitions that lead to multibillion-dollar deals. It’s at the base of everything. In recent years, Goldman has launched a new data streaming business for the bank’s hedge fund and wealth management clients.
For decades, banks have taken the lead in data collection, amassing large amounts of information about transactions, transactions, companies, loans, etc., among others. However, little thought was given to how that information was captured and used to compare with other information hidden elsewhere in the company.
As Chief Data Officer, Rafael’s job is to organize all of Goldman’s data and make it easy to search and collaborate on. Doing so is paramount to creating the most effective AI models. As Goldman focuses its efforts on AI, Rafael has emerged as a key player influencing the bank’s future.
He details to BI how he organizes a team of around 500 people to get the most out of data, and how the data function is integrated with the rest of the bank. Did. As a top data executive responsible for both strategy and engineering, Rafael faces a high-stakes crossroads for his career and for Goldman.
How Raphael’s data platform team is leading Goldman’s AI innovation
Raphael’s day can vary greatly, but the beginning is always the same. A relatively new father, he wakes up at 6:30, plays with his 2-year-old for 30 minutes, then goes to work. Arriving at 200 West Street, Rafael’s day jumps between several main focuses.
One is the data platform team, which oversees the entire data stack, including the infrastructure and how people access data. The team is responsible for organizing, standardizing, and cleaning data. An example is ensuring that the word management is consistently spelled MGT rather than “MGMT” or “management.” Their goal is to make it easier for end users to find data and work with the same dataset in an organized way.
Cleaning up this data helps companies experiment with AI. Goldman uses her AI to summarize and extract information from documents related to loans, mortgages, derivatives, and more. AI is not new to Goldman. He founded the bank’s machine learning and AI team in 2018. But with recent advances in AI allowing new content to be generated, the bank is using its AI to enable its engineers to analyze other people’s code, and even non-technical employees. I was able to entrust more work to them. Comes with software.
Goldman isn’t the only company with AI in its brain. Wall Street companies are eager to leverage his AI to gain an edge in everything from back-office efficiency to investing and trading. The AI craze has reignited a waning talent market, with some banks hiring hundreds of technologists to execute their AI plans. Meanwhile, other Wall Street firms such as JPMorgan have launched new divisions led by data, analytics and AI.
Despite all the talk about AI on Wall Street, Rafael calls it a tool that sits on top of data, “just like business intelligence was five years ago and data science was before that.” He said he thought it was just a matter of. “All of these things are built on this foundational layer that we’re creating,” he told BI over Zoom, wearing a blue plaid shirt and slicked back hair. Told.
How Goldman connects data and business
Rafael has worn many hats throughout his tenure at Goldman. He started his career at the bank as an analyst in the technology sector in 2003 after he graduated from the University of California, Berkeley. He has risen through the ranks leading various engineering teams, including an engineering SWAT team that works directly with his CTO on special projects. He was also part of the team that built SecDB, one of Wall Street’s first enterprise-wide securities pricing and risk management systems. He was named Managing Director in 2013 and became Partner in 2020. Four years ago, he was named Goldman’s sole head of data and head of data engineering.
Inspired by his work at SecDB, Raphael started a data design and curation team. This task force is made up of smart analytics gurus across technology, data and finance who help other teams tackle tough data problems and build bespoke data solutions. The data design team’s day-to-day work is constantly changing to solve business problems, he said.
“One day we’ll be talking to an asset manager and saying, ‘What are we going to do with our private fund data? We want more data to enrich it,'” Rafael said. Another day, sales and trading could suggest new ways to analyze Q10 filings using alternative data to help customers generate new insights to make better deals. He said that there is.
In late 2020, Mr. Rafael brought some of Goldman’s alternative data processes under his wing, allowing trading desks to obtain proprietary data more quickly.
So-called alternative data includes non-traditional information used in decision-making and investing, such as what individual traders say on Reddit message boards. This information is at the heart of some trading desks’ secret sauce.
Rafael said Goldman always has a centralized data sourcing team that acts as a “gatekeeper” and signs vendor agreements. But the engineering part, such as ensuring the quality of the data and entering it properly into the database, was previously done by a separate team that outsourced the data, he said. With Rafael’s team handling the data engineering portion, trading desks will be able to obtain data more quickly and Goldman will be able to connect the dots between alternative data and provide curated insights to trading desks. added.
“They love the speed, but also the kind of curation of light that we do. That was very powerful for them,” he said.
Success as a data scientist
In addition to these focuses, the Chief Data Officer ensures that the bank manages critical governance functions and manages the most critical information streams used across the company, including real-time and historical market data from exchanges. It also helps supervise some of the.
After all, Raphael ends his day the same way he started it. That means spending time with my two-year-old.
Raphael offered two pieces of advice to young data engineers looking to advance on Wall Street.
Solve problems by first understanding your data and its context, understanding how it connects to the world and your domain. Then, understand the business, he said.
“Focusing on that area and marrying it with data, data expertise, is what will get you to a kind of unicorn status as a data person,” he said.
