projects.
research.
We provide students from all technical backgrounds with the opportunity to learn and apply various techniques, packages, and concepts in data science. Through a hands-on project with one to two other teammates, you will go through the entire data science life cycle to create a product that impacts daily lives in social good sectors and bridges the gap between technical and practical knowledge.
industry.
We partner with nonprofit organizations, medical centers, research institutes, progress-driven companies, and more to deliver technical projects that help them better serve their communities. With one or two Project Managers and four other teammates, you will sharpen your skills and be able to impact organizations within 10 weeks.
the right fit.
When joining a Research or Industry project, the main difference boils down to the required experience level and whether there is an external organization involved. Each project is supported by the DataGood team and allows members the same levels of mentorship, guidance, and growth catered to their learning style. We encourage our members to survey the project offerings in both Research and Industry areas each semester and apply to those in areas where they would like to develop and sharpen their skills in.
fall 2025 industry projects
Synopsys
We're harnessing advanced ML pipelines to revolutionize chip design by predicting power use early, eliminating slow, energy-intensive simulations. We're accelerating chip design while cutting computational waste and driving greener, more sustainable hardware engineering. By partnering with Synopsys, our project helps pioneer energy efficiency, while setting a new standard for responsible semiconductor innovation.
Teach for America
We're partnering with Teach for America to explore how to diversify the teacher workforce and close representation gaps in classrooms nationwide. By spotlighting BIPOC leadership within TFA and elevating stories of teachers of color, we're helping pave the way toward more equitable classrooms. Our work empowers education leaders with data-driven insights to reimagine representation and bring lasting impact to schools across the country.
The Nature Conservancy
We're developing a machine learning-based snowpack modeling project focused on Arizona forests, combining LiDAR forest data with climate datasets! We'll deliver practical, data-driven insights to guide forest and water management efforts. We want to help spread awareness and inform how different forest treatments impact snowpack across different regions and conditions, and better understand water resilience.
Stride
Our project involves building an AI-powered text and emoji analytics framework that makes online communities safer, more inclusive, and more engaging. We're helping brands, students, and organizations foster healthier digital interactions through our actionable insights!
fall 2025 research projects
Modeling Behavioral Biases in E-Commerce
This project investigates behavioral biases in online shopping by analyzing how factors such as product reviews, pricing strategies, and scarcity cues influence consumer purchase decisions. The team conducts exploratory data analysis (EDA) and statistical tests to quantify the effects of these biases, identifying which factors most strongly shape purchase likelihood and consumer decision-making. Predictive models, such as regressions, are developed to forecast purchasing behavior, and findings are visualized through a dashboard highlighting key behavioral insights.
AI for Smarter Healthcare
This project develops an intelligent healthcare resource management system designed to forecast hospital capacity demands and optimize resource allocation. By leveraging predictive modeling and optimization techniques, the project supports healthcare providers in responding effectively to surges in patient needs, ultimately enhancing patient care and operational efficiency. The work encompasses the full data science pipeline, including data preprocessing, exploratory data analysis (EDA), and machine learning modeling using libraries such as PyTorch.
Urban Health and Predictive Analysis
This project explores the relationship between urban walkability and mental health outcomes across the United States. Using data from the EPA Walkability Index and CDC PLACES Project, the team examines whether cities with more walkable neighborhoods exhibit lower rates of poor mental health. The project involves data cleaning and transformation, exploratory data analysis (EDA), and the construction of predictive models to estimate mental health outcomes at the census tract level. Results are presented through an interactive dashboard built with industry-standard visualization tools such as Tableau.
spring 2025 industry projects
HealthTrust
The HealthTrust project focused on building a data-driven risk assessment tool to evaluate the operational health of Bay Area nonprofits. Using survey data, the team identified key risk categories including leadership burnout, financial instability, and compliance challenges. A scoring model was developed to quantify organizational risk and help prioritize technical assistance and funding. Tableau dashboards and data visualizations were created to synthesize findings for HealthTrust’s internal strategy and grantee engagement. The final deliverable provides actionable insights to support capacity-building across the nonprofit ecosystem.
Tilli
The Tilli project aimed to develop an AI-powered tool that digitizes and grades handwritten K–5 math exams, with feedback delivered via WhatsApp. The team experimented with both public and HuggingFace OCR/LLM models to evaluate their performance on diverse handwriting styles. Key challenges included interpreting multilingual, messy handwriting and generating clear, SEL-aligned feedback. The final recommendation was a hybrid model strategy combining fast OCR with deep reasoning models like Claude and GPT. The deliverable provides a foundation for Tilli’s MVP, with a focus on scalability, low-bandwidth environments, and personalized learning insights.
Team Rubicon
The Team Rubicon project focused on developing a predictive and decision-support tool that identifies and prioritizes communities most vulnerable to extreme fire and flood events. This tool will enable more proactive and efficient deployment of mitigation services (e.g., brush removal, hazard home ignition zone assessments, sandbagging, etc.) by integrating publicly available data, historical event records, and county-level emergency services information. Ultimately, this project aims to improve the impact of Team Rubicon’s volunteer efforts and resource allocation by providing data-driven recommendations on where and how to intervene.
Noora Health
The project aims to develop a voice data pipeline for Noora Health, enabling automated ingestion of voice recordings, exploratory data analysis (EDA), and multi-language transcription. The DataGood team will build an ingestion pipeline to retrieve and store media files from turn.io and exotel, conduct EDA to extract key insights and patterns from the data, and implement a transcription pipeline to convert voice recordings into text across multiple languages.
Ashby Village
DataGood’s partnership with Ashby Village (including Ashby Village’s Data Group) focused on improving event attendance tracking and creating a user-friendly dashboard for visualizing service utilization and event attendance data. This project aims to streamline attendance data entry for those who attend Ashby Village in-person events and provide insightful data visualizations for staff, board members, and volunteers.
spring 2025 research projects
Heart Attack Analysis & Prediction
Heart disease is the leading cause of death in the developed world, making early prediction and prevention crucial in reducing mortality rates. This project focuses on analyzing heart disease risk factors and developing a machine learning model to predict the likelihood of heart attacks. Using a dataset from Kaggle, we will apply data preprocessing, exploratory data analysis (EDA), and machine learning techniques to identify high-risk patients based on key health indicators. For the data analysis portion, we will leverage Data100 methodologies and industry-recognized libraries such as Pandas, Matplotlib, Seaborn, and more to explore trends and correlations within the dataset. Our machine learning pipeline will experiment with various models, including Logistic Regression, Random Forest, XGBoost, and other classification algorithms, to evaluate performance and optimize predictions.
Food Desert Prediction
Food deserts are areas where residents have limited access to affordable and nutritious food. Millions of households are impacted by food insecurity globally, especially in underserved or rural areas. Identifying and predicting food deserts can help policymakers allocate resources effectively, for example by opening grocery stores or providing food subsidies such as SNAP/EBT. This project aims to use K-Means clustering for geographical analysis and a Random Forest Classifier with feature engineering to uncover patterns in the data and predict food deserts.
Wildfire Impact and Resource Allocation
The mission of this project is to analyze social media data in California to identify trends using natural language processing (NLP). The goal is to assess personal needs and assist users affected by fires through data science techniques. This project will cover exploratory data analysis and visualizations to effectively present findings. Research members will learn to use the Twitter API to collect text data and apply various NLP techniques to gain insights into resource allocation and user impact. By the end of the project, research members will create and present a dashboard that showcases key entities and insights from sentiment analysis related to this research.
fall 2024 industry projects
Oakland Promise
Oakland Promise empowers Oakland students through offering financial resources, educational programming, mentorship, and a supportive community. The main goal of this project is to determine metrics to assess college affordability, develop survey questions and conceptual framework to measure the impact of Oakland Promise’s College Access scholarships on college affordability, administer the survey to current scholars, and analyze the results.
Point Blue Conservation Science
Point Blue is a nonprofit organization focused on climate-smart conservation to help ensure that wildlife and communities continue to thrive in decades to come. As leaders and innovators in conservation science, Point Blue is filled with scientists, educators, and habitat restoration specialists. Data Good’s collaboration with Point Blue focuses on the Antarctic Wildlife Program in which we use data driven tools to access Adélie penguin ecology. This project is mainly focused on computer vision models.
Tenants Together
Tenants Together is a non profit organization based in San Francisco, CA that is dedicated to defending and advancing the rights of California tenants to safe, decent and affordable housing. The purpose of this project is to analyze Tenants Together’s data as well as publicly available data (ie. census data) to understand the key metrics that make up ‘what an average Californian tenant’ looks like. We will do data analysis as well as machine learning such as regression analysis and causal inference in this project.
Think of Us
Think of Us is a nonprofit organization dedicated to transforming the nation’s child welfare system to improve the lives of millions of children and families each year. Our collaboration with Think of Us focuses on using natural language processing and data analytics to analyze and enhance the efficiency of their operations to help them achieve their mission within the child welfare sector.
fall 2024 research projects
Bias in Predictive Policing
With data science becoming a large part of policing in America in the form of predictive policing, it is imperative to ensure that these models are not discriminative against certain groups. Recent analyses, however, show that such policing models can be harmful to the criminal justice system. This project will teach and utilize prevalent data analysis techniques, such as data cleaning and building data models, to create and evaluate models for predictive policing based on the COMPAS algorithm. Finally, an interactive Bokeh dashboard will be incorporated to demonstrate the findings of the research analysis.
Financial Vulnerability Assessment
The goal of the Financial Vulnerability Assessment project is to create a KNN (K-Nearest Neighbors) model capable of forecasting high and low financial risks based on demographic factors. By leveraging machine learning techniques, the project aims to educate the general public on insights into their financial vulnerabilities, thereby fostering a more informed approach to managing finances. By the end of this project, you will provide an algorithm that uses personalized risk assessments to ultimately enable better decision-making in the face of economic challenges.
Data-Driven Insights(Returning DataGood Members Only)
The mission of the Intermediate Track is to bridge the gap between Research and Industry projects by providing developers with a controlled environment to learn the soft and professional skills they need to succeed in data consulting. The Intermediate Track combines a preliminary curriculum on strategy, communications, and essential data tools with a fabricated client scenario to prepare developers for working with real clients in Industry, and beyond.
spring 2024 industry projects
Healthy Planet Project
This project will Use data from the Healthy Planet Project to analyze the most immediate needs in SF to improve the environmental health of the city and investigate what environmental causes SF residents donate to the most.
Multicultural Institute
Multicultural institute is a non-profit dedicated to empowering immigrants, especially those in the Latinx community. This project will use data analysis to identify areas for program improvement across their 3 chapters. Additionally, this program aims to showcase their impact and advocate for their support of immigrant communities.
spring 2024 research projects
Sustianable Space Debris Management
The growing concern over space debris has significant commercial implications, as the sustainability of space activities is crucial for both governmental and private sector stakeholders. With over 23,000 tracked objects 10cm or greater as of July 2023, and the incident in August 2023 where space debris came close to the International Space Station, there is an increasing recognition of the need for active debris removal. The Kessler Syndrome, which describes a cascade of collisions due to high object density in low Earth orbit, poses a direct threat to the infrastructure of space agencies like NASA, SpaceX, and the European Space Agency, both of whom are actively seeking solutions to combat this problem. The potential for commercial ventures in this area is substantial, as these agencies are looking to develop technologies for debris monitoring, mitigation, and removal.
DetectAI
The goal of this project is to harness the power of Convolutional Neural Networks (CNNs) to revolutionize breast cancer detection. Breast cancer is a significant global health concern, and early diagnosis plays a pivotal role in improving patient outcomes. Our project aims to create a robust and sophisticated CNN model designed to analyze medical imaging data, specifically mammograms, with unparalleled accuracy.
Forest Fire Prediction
The ongoing rise of global warming and climate change have led to the increased need for more accurate and efficient forest fire detection. Using real-world data, this project will teach members fundamental data science and machine learning skills to analyze the predictability of forest fires. Members will have a chance to choose their attributes to look into to build their own machine learning model about the predictability of fires in a certain area, learning how to help minimize the environmental and humanitarian risks of forest fires.
fall 2023 industry projects
Scientific Adventures for Girls
Using the Scientific Adventures for Girls' after-school program student and parent data from the last 8 years to create a dashboard and infographics for their website and marketing resources.
Young Women in Bio
By manipulating YWIB's national datasets in numerous ways, we will examine the factors that lead to more young women entering STEM, possibly how these factors can be increased, creating deliverables that help YWIB find the best locations to market to, and ways they can improve their workshops.
Roots of Success
This project aims to help Roots of Success better understand and communicate their impact by analyzing and visualizing data resulting from their environmental education programs, enhancing their ability to attract partners and donors, and set the organization up for continued success in the future.
fall 2023 research projects
Real or Fake?
Through this project, students will learn the necessary data science fundamentals to distinguish between fake and real news by building and testing their own classifier model to evaluate the credibility of information sources. Students will have opportunities to explore topics such as feature engineering and machine learning while conducting their own exploratory data analysis.
Gun Violence & Gun Control
This project empowers students to deeply research and analyze a chosen subtopic within gun violence and control. They can explore areas such as violence trends, impact of control measures, predictive modeling, community engagement, firearm accessibility, victim support, and illegal firearm activities.
Sustain AI - Sorting Trash for a Greener Future
Students will explore how we can utilize cutting-edge models to sort waste by building their own waste classification machine learning model, thus optimizing recycling processes, minimizing landfill waste, and mitigating ecological harm.
An Ideal Trip
Students will create a data model that makes planning weekend trips for college students seamless and stress-free. Based on their weather preferences, number of restaurants, and other metrics, this model will plan their ideal trip location and itinerary. Students will be able to learn how to use multiple linear regression in R, and walk away with a better understanding of how they can use their data science skills in everyday life.
spring 2023 industry projects
League of Women Voters (LWV)
Using the League of Women Voters' membership data, the team will be analyzing their performance in membership growth and community outreach. This will better guide their decision-making process of where to strengthen their efforts and allocate resources. Students will also be utilizing predictive models to project the organization’s growth give them metrics to measure their impact.
Berkeley Disaster Preparedness Neighborhood Network
Using BDPNN's member list datasets and disaster preparedness zoning, we'll perform data analysis to develop recommendations for helping BDPNN better mobilize volunteers and determine regions to improve deployment for responding to natural disasters. Through overlaying the member data with demographic layers and zoning data, we’ll categorize members into zones for emergency responses. Additionally, we'll analyze disasters and predict the relative risk.
spring 2023 research projects
Analysis of Education Equality and Income
In this project, students will use various datasets, like the US census, that contain information about education quality as well as income rate of American households in certain school districts. The team will use our findings to analyze the correlation between the quality of education and household income rates. Using those same findings students will be able to work together and create a project that will give us an opportunity to share with our peers any conclusions we have made based off the data.
Analysis of Pandemic and Unemployment
This project will use datasets on COVID-19 and unemployment rates in the United States to take a look at how different industries within the economy have responded to the economic decline caused by Covid-19, as well as understanding any patterns that may emerge. Members will ultimately create a final deliverable with data visualizations that detail any conclusions they may draw on the topic from their project.
Cancer Detection with Data Analysis
In this project, students will use healthcare datasets, such as cancer patients data, to examine the lifestyles of cancer patients, and members will be taking a look at how certain factors cause people to be more susceptible to cancer. Students will also explore the relationships between the characteristics and traits of individuals in different physical environments. This will culminate in members creating a project that draws patterns and reaches conclusions related to the topic at hand.
fall 2022 industry projects
Community Resources for Science (CRS)
The main goal of this project is to help the Community Resources for Science (CRS) analyze survey data from 10+ years of surveys stored in CSV files, possibly selecting only certain years to look at. CRS anticipates looking at 6-8 key questions like teachers who are newer vs more experienced; teachers from 3 different school districts, teachers' level of confidence and amount of time spent on teaching science, etc. The primary reason for looking more into these key questions is to help CRS better understand the trends and differentiations among these key questions.
Food Bank of Contra Costa & Solano
The main goal for this project is to help the Food Bank of Contra Costa & Solano better understand the communities they are serving. In order to accomplish this, we will add layers to the existing Tableau dashboard that is currently being used to map and visualize food distribution and needs. Adding more demographic layers to the map will allow the Food Bank of Contra Costa & Solano to better understand what specific communities they are reaching. Additionally, we would help analyze impact through more detailed visualizations in a dashboard.
fall 2022 research projects
Tracking Causes of Climate Change
This project focuses on using data science methods such as data visualization, regression, and classification to identify and predict fraudulent entries in healthcare provider data. Suggested prereq: Data 8
COVID sentiment analysis using NLP
As we move into a world that has seen the numerous waves of COVID-19, we want to learn more about how people have adjusted to the global pandemic we are enduring. Topics of interest include: mental health, healthcare, politics, vaccines, employment etc.
Unemployment Rate Analysis
The goal of this project is to utilize multiple datasets to learn important techniques in data science, and educate students on how to provide clear, presentable data. The final project will use datasets on the general US unemployment rate from 1948 to 2022 and more detailed percentages classified based on age, race, education level.
Gender Income Inequality
The goal of this project is to utilize multiple datasets to learn important techniques in data science, and educate students on how to provide clear, presentable data. The final project will use datasets on the income of men and women in various tech jobs where members will be taking a look at how various factors such as seniority, level of education and age can contribute to gender inequality in income.
spring 2022 industry projects
Interactive Map to Detect Environmental Inequities in Communities
This project is an interactive map that layers data about demographics and environmental issues to detect inequities between communities.
Anomaly Detection in Healthcare Providers
This project focuses on utilizing data science techniques such as data visualization and exploratory data analysis to find trends in climate data. We will examine the factors which affect the Earth’s temperature in various locations.
NMDC
This project will contribute to the NMDC's sample-annotator API for repairing biosample metadata that doesn't adhere to standards like MIxS. The repaired data will drive collaboration and open science.
spring 2022 research projects
Climate Change Data Analysis
This project aims to create a model to predict future CO2 levels globally, or other indicators of climate change, while reviewing important packages and tools for analyzing and visualizing data. Suggested prereq: Data 8
Online Learning Effect on Student Performance
This project aims to understand the effects of online learning on student performance. Using important data science techniques such as regression, we will look at how various factors affect student performance in both online and in-person learning environments. Suggested prereq: Data 8
Tracking COVID-19 Vaccine Equity
In this project, we aim to employ national vaccination data to analyze COVID-19 vaccination programs in different countries. We would like to answer important questions about how the world is dealing with the surging virus. Do richer and poorer countries have very differing access to COVID-19 vaccines? Which country has the most efficient vaccination program? Does the type of vaccine being used matter? Or are there any geopolitical factors in play? Thanks to the comprehensive dataset from Our World in Data (OWID), there’s plenty of room for us to track COVID-19 vaccine equity across the globe. Difficulty level: Easy to medium. Will cover the basics of everything. No prior technical experience required.
fall 2021 industry projects
Student Classifier
This project is a web application that Downtown College Prep can utilize to improve English learner reclassification and cohort identification for their students.
Ocean Trash Dashboard
This project is a live dashboard with Pacifica Beach Coalition clean up data that’s deployed online for progress tracking and education.
fall 2021 research projects
Bay Area Public and Environmental Health Analysis
This project aims to introduce a variety of packages essential to analyzing and visualizing data, as well as exploring different sectors of social good.
COVID-19 Sentiment Analysis
This project aims to use sentiment analysis on tweets about the COVID-19 vaccination to understand how people feel about it across the world.
World Happiness Report
This project is to understand data about the World Happiness Report and what factors impact global wellbeing.
spring 2021 industry projects
Firewatch Chatbot
The Firewatch Chatbot aims to: provide reliable info on fires close by, answer fire preparedness questions, and aid CalSAFE’s donor engagement and outreach
Transfer Student Classification
Building an NLU topic model using over 10 years of scholarship submissions to both identify the problems students speak about in an unsupervised manner and analyze growth and evolution of them in time-series text data.
spring 2021 research projects
DC + Chicago Crime Analysis
Leveraging DC + Chicago crime data to visualize trends across factors such as race, socioeconomic status using regression and geospatial analysis
Natural Language Processing on US Legislature
This project aims to analyze Pennsylvania legislature using NLP techniques such as TF-IDF, Part of Speech tagging, and more
Cardiovascular Disease Classifer
Building a classifer to predict various severities of cardiovascular disease in patients
past projects
Nehemiah Gateway USA
Provide actionable information that will increase donor reach and engagement using data science and NLP
California Institute For Energy And Environment
Exploring racial biases/implications in solar energy and equity with respect to credit score data
Medical Insurance Analysis and Classifer
Created data science project that utilized a Python GUI to visualize different data science techniques on insurance health charges data using Pandas, Sklearn, and Tkinter
Mapping for Environmental Justice
Mapping for Environmental Justice (MEJ) is an initiative to create interactive and publicly-accessible maps displaying environmental justice data for individual states. With guidance from the residents of impacted communities, MEJ combines environmental, public health, and demographic data into an indicator of vulnerability for communities in every state.
Electronic Pathology Report Classification
In partnership with the Cancer Registry of Greater California, this project uses ML and NLP on electronic pathology records to identify reports missing critical diagnostic information especially for ethinic minorities in the state of California.
