Zicklin Grad Students Unlock New York’s Immigrant Stories with Data Analytics

New York is a city of immigrants. Year after year they arrive by the thousands, each with his or her own unique story of hardship, hope, and survival.

But what if there were a way to unlock new insights into the immigrant experience by using AI? That’s what a team of Zicklin School of Business graduate students in Joshua Moritz’s marketing analytics classes did over the past year.

Clockwise from upper left: Shrusti Bhalala, Hesham Ebrahim, Nikita Jain (all MS Marketing students), Lecturer Josh Moritz, MS Marketing students Nataly Olarte, Monica Phan, and Shannon Thomas

In a collaboration of the Zicklin School with Baruch’s Sandra Wasserman Jewish Studies Center (which provided a microgrant) and the Tenement Museum on the Lower East Side, Moritz’s students used various generative AI tools—Claude, ChatGPT, and CoPilot—to build a large language model (LLM) that analyzed transcripts of hours of interviews with former residents of the neighborhood, to identify key themes and reveal patterns that weren’t apparent from reading transcripts manually.

Using the LLM, students analyzed 22 first-person interviews that had been digitized from more than 35 hours of taped interviews with merchants from the Lower East Side who were of Chinese, Jewish, Irish, Latino, and Italian descent. The interviews had been recorded from the 1970s through the 1990s for a public access cable TV show created by Bea Moss. (Moss was Moritz’s aunt, and he gained access to her recordings after she died.)

Although some of the merchants were not immigrants themselves, all had a front-row seat to the demographic changes in the neighborhood as longtime Jewish and Italian residents made way for growing numbers of Puerto Ricans, Chinese, and other groups.

The LLM tools helped contextualize and chronologize this dense neighborhood characterized by massive waves of immigration: In a boon to local historians, they automatically extracted dates, names of people, places, and organizations. They also analyzed emotional content, scoring interviewees’ sentences as positive, negative, or neutral, thereby unearthing striking similarities despite cultural and linguistic differences.

“We found that when different communities narrate the same experience, there is a universal immigrant grammar despite mother tongues being significantly different from one another,” noted Zicklin student Shannon Thomas (MS, ’27), who provided presentation and documentation support for the project.

The dominant sentiment, observed in a third of interviews, was “nostalgia and longing for a world they once knew,” Shannon added. The second-highest sentiment observed in interviews was “pride” in triumphing over hardship, said project manager Nikita Jain (MS, ’26). “Hardship/sorrow” was the third-highest sentiment, and the fourth highest was “hope and optimism.”

Working with the Wasserman Jewish Studies Center, students presented their research as part of a conference on antisemitism. Moritz’s student Hesham Ebrahim (MS, ’27) created “word clouds” from the most common words repeated in Jewish newcomers’ narratives —“immigration,” “journey,” “hope,” “survival,” “work,” and “community.”

Hesham said he was particularly moved by an interviewee’s account of how lighting candles on the Sabbath “transformed cramped rooms into sanctuaries” and that the experience led him to reflect on his own journey as a voluntary immigrant from Egypt. “I also learned about the origins of New York foods I had never encountered before, like bagels and babka,” he added.

Shannon was intrigued by a detail in an interview she transcribed from 1930, in which the interviewee recalled that when the Empire State Building was being constructed, his older brother would carry a gallon of ice cream up 56 stories and sell it to the construction workers. “He’d make two dollars on a gallon and that fed their family of six,” she said. “In 2026, two dollars can’t even get you on the subway.”

Lecturer Joshua Moritz (Allen G. Aaronson Department of Marketing and International Business) said that the transcripts provided an unexpected opportunity to apply LLM class learnings to work in a real-life situation, giving voice to the merchants whose insights might have been passed over for generations. While not a business problem that Zicklin students would normally focus on, the project provided them with an excellent opportunity to demonstrate their skills in developing LLM models using real interviews, an experience they can use with future employers.

For more about graduate programs at the Zicklin School, click here.

Categories: Alumni News, Faculty, News, Student News