Data for Good: Bloomberg supports data scientists’ work with nonprofits and municipalities to solve real-world problems

Data scientists are working alongside local, national and international organizations.

Special Video: 2017 Immersion Day @ Grand Central Partnership, premiered at Bloomberg’s 2017 Data for Good Exchange, September 24th, 2017

In 2016, Bloomberg and NYC Media Lab collaborated on the initial Bloomberg Data for Good Exchange Immersion Day program, in which data scientists from NYC Media Lab’s consortium of universities were paired with nonprofit organizations and municipalities to assess real-world data. Taking inspiration from its success, Bloomberg’s Office of the CTO launched this year’s cohort to further explore how data science research in the City’s universities can be used to impact the public sector.

The Immersion Day program coincides with Bloomberg’s annual Data for Good Exchange (D4GX). At the conference, which took place on Sunday, September 24th at Bloomberg’s Global Headquarters in New York City, leaders in technology, policy, and governance convened to discuss how applications of data science can improve public service. Participating Immersion Day data scientists and leaders from their host organizations sat down with NYC Media Lab Executive Director Justin Hendrix to discuss their project and lesson’s learned.

2017 saw eight data scientists selected by Bloomberg to participate in the program, with a focus on governance and community-driven missions in major cities in the U.S. and overseas: New York, Philadelphia, Paris and Bogotá, Colombia. The researchers spent several days this summer working with their host organizations to execute a data science project based on specific needs, goals and challenges. Host organizations spanned from government offices to nonprofits in areas like education, technology and healthcare.

Across the private sector, integrating data science into corporate culture has helped teams provide better service and improve internal leadership. A core objective of the Immersion Day program is to aid the public sector in exploring similar insights, as many organizations have fewer resources and limited access to the most current data science research. What kind of data is relevant to a civic organization? How can the academic community help create a set of suggestions for governments and nonprofits? And, most importantly, how can data science help strengthen a social mission? These types of questions served as the starting point for the researchers and their host organizations as they collaborated throughout the program. In exchange for giving access to their unique data sets, civic and nonprofit hosts are provided with technical assistance and expertise from the data scientists.

NYC Media Lab’s Justin Hendrix moderating the Bloomberg Immersion Day panel at Bloomberg’s Data for Good Exchange 2017.

Over the course of the program, the collaborations reflected great diversity in terms of needs and objectives. Working together with Bloomberg Associates, some of the data scientists aimed to help city government successfully tackle complex and difficult challenges to positively impact the quality of life of their citizens. For example, the team in Paris worked to create a new dashboard to better understand how key performance metrics are achieved, with the goal of helping the city with its plans to launch new, publicly-supported civic spaces for startups. In Bogotá, they created a bridge between citizens and the city’s administration, helping directors and heads of citizen services better understand and visualize complaints from residents.

Some of the other data scientists partnered with nonprofits in the U.S. to help them find new ways to leverage their existing data to improve their delivery of services. For example, outside Philadelphia, healthcare nonprofit Bringing Hope Home, which works to provide assistance to families with cancer, explored how to use data science to shorten the wait time for services. And three of the projects were based right here in New York. One researcher worked with continuing education nonprofit Matriculate to build a real-time data platform to better communicate with high achieving high school students from low income communities. Another worked with Grand Central Partnership, one of the largest business improvement districts in the world, better understand how consumers interact with various assets (e.g., benches, light poles, trash cans, etc.) in this busy portion of midtown Manhattan to better ensure their upkeep. And lastly, a team at iMentor, a nonprofit that pairs industry executives with high school students for professional training, built a neural network to weigh performance evaluation data from mentors and determine how likely their student participants are to find career success.

Bloomberg Immersion Day data scientists and their hosts convene at Bloomberg’s Data for Good Exchange 2017.

In a short amount of time, host organizations were able to receive a broad look at emerging trends in computer science, engineering and information systems. By encouraging interdisciplinary thinking and rapid experimentation, host organizations learn what’s possible — perhaps helping them to continue re-strategize their upcoming roadmaps to embrace newer, more advanced technologies.

Below, you’ll find details about the fellows selected by Bloomberg, including their university affiliation and host organization.

Meet the 2017 Immersion Day cohort!


Yuan Lai

Working with: Grand Central Partnership, New York City | website

Lai (NYU Tandon + CUSP) is a PhD candidate in urban systems at NYU Tandon School of Engineering. He holds a M.S. in urban informatics from NYU Center for Urban Science and Progress (CUSP) and a master’s in urban planning. Yuan’s research focuses on informatics and data science for urban development integrating data analytics, physical computing, and system engineering.

Bhavya Ghai

Working with: Matriculate, New York City| website

Ghai is a PhD Student in the Computer Science Department at Stony Brook University, specializing in Data Science & Visualization. His recent work deals with visualizing high-dimensional, multivariate data into lower dimensions using genetic algorithms. As a Data Science for Social Good fellow at Georgia Tech, he worked on housing justice and policy analysis. He also worked as a research intern at Indian Institute of Technology, Delhi where he tried to predict solar radiation to ensure power supply-demand equilibrium. Previously, he completed his Bachelor’s & Masters in Information Technology from Indian Institute of Information Technology, Gwalior, where he worked on ensemble learning and robotics. He has received many awards for data science research, including the Chairman Fellowship Award.

Geoff Perrin

Working With: Bloomberg Associates, City of Bogota

Perrin is a recent recipient of a master’s degree in Urban Informatics from NYU’s Center for Urban Science and Progress. He has been involved in data science projects ranging from predicting the occupancy of homes in Detroit, to predicting lot level solid waste generation in New York City, to classifying bike lane and street quality using accelerometer and imagery data collected from cell phones. While at NYU, he was chosen to work as a Graduate Research MacArthur Fellow under Dr. Constantine Kontokosta.

Pascal Luttgens

Working With: Bloomberg Associates, City of Paris

Pascal is a junior data scientist at OpenDataSoft. His current mission consists of analyzing and monitoring log data to better understand user behaviors as well as setting up new data pipelines to allow for producing new metrics and indicators to be exploited. In his free time, he likes to dig into specific fields of machine learning to broaden his knowledge, knowing that it will eventually be useful for future challenges. For example, he is currently trying to make a robust AI for games using reinforcement learning and Q learning.

Xavier Gonzalez

Working with: Bloomberg Associates, City of Bogota

González is a Data Scientist with +15 years of experience in the Industry and Academia. Before his graduation in Industrial Engineering, he started working for big companies like IBM and General Motors. Since then, he has gained business experience in diverse areas such as Operations, Marketing, and Strategic Sales Planning. After a 2-year assignment to General Motors do Brazil, he became professor of Operations Research in the public University of Buenos Aires, where he also conducts research. His contributions in the fields of Decision Making, Optimization and Agent Based Modeling extend to +20 papers and presentations. He received the ‘Argentine Presidential Fellowship Award’ to pursue a MS degree in Data Science in Columbia University and the Peruilh Scholarship Award to fund his PhD studies. During his stay in NYC, his team’s projects won the first prize in the ‘MINDS Innovation Challenge’ and the ‘Microsoft prize’ in the DEVFEST Hackathon.

Vincent Major

Working with: Bringing Hope Home, Malvern (Greater Philadelphia Area)| website

Major is a 3rd year PhD student at New York University’s School of Medicine working in Medical Informatics. More specifically, he is applying machine learning techniques to electronic medical record data for a variety of tasks. Originally, he was trained as an engineer with a BE(Hons) in Mechanical Engineering and ME in Bioengineering from University of Canterbury, New Zealand. His research in New Zealand focused on mathematical modeling of respiratory dynamics for critical care patients supported by invasive mechanical ventilation.

Evgeny Bagdasaryan

Working with: iMentor, New York City | Website

Bagdasaryan is a 2nd year Computer Science PhD student at Cornell Tech, where he works on recommendation systems and intelligent assistants. He focuses on applying Deep Learning methods towards heterogenous data to derive useful recommendations and support decision making. Eugene graduated from Bauman Moscow State Technical University and worked for Cisco Systems R&D team in Moscow, Russia for 2 years.

Michael Sobolev

Working with: iMentor, New York City | Website

Michael is a post-doctoral fellow in the small data lab at Cornell Tech in New York City. Michael recently completed his PhD in behavioral economics from the Technion in Israel. Currently he focuses on connecting behavioral science and technology to study and improve human behavior in the areas of productivity and health. He also tries to improve recommendation systems and machine learning by brining insights from the study of human behavior.


To learn more about this year’s program, contact NYC Media Lab’s Manager of Partnerships Amy Chen (amy.chen@nycmedialab.org), or Bloomberg’s Head of Innovation Communication Chaim Haas (chaas30@bloomberg.net).