Do you have a complex data challenge that could benefit from a dedicated team of master’s-level data scientists? The UVA School of Data Science welcomes project proposals from organizations across industry, government, academia, and the nonprofit sector.

Capstone projects challenge UVA data science students to explore, integrate, and analyze data to solve real-world problems. Project teams consist of three to five students and one or more faculty mentors. These mentors work with the sponsoring organization to ensure the project meets the sponsor's needs as well as the educational goals of the program.  

All members of the project teams will have completed courses in computational methods for data science, exploratory data analysis, and programming in R, Python, and SQL. During the program, students also learn feature engineering, data cleaning and wrangling, agile software development, data ethics, project management, team building, technical writing, and oral communication. 

Application Close Icon Close

If you are interested in sponsoring a capstone but still have questions, connect with director Adam Tashman at apt4c@virginia.edu. If you’re ready to submit an application, download the template and email to DataCapstones@virginia.edu. Please keep in mind that applying is just the first step to further discussion and refinement. We approach the sponsorship process as an open dialogue to ensure the capstone is successful for the sponsor and our student teams. 

Sample Agreement Close Icon Close

Download the sample agreement between a sponsor and the UVA School of Data Science.

What Makes a Successful Project Close Icon Close

Project Scope

Satisfying projects are adequately specific to allow students to deliver in the time frame, but also open-ended enough to allow for creativity and problem-solving. It can work well to provide focused goals and reach goals.

The pillars of data science are data acquisition, data engineering (e.g., data integration, wrangling), data analysis (including machine learning), and data communication (e.g., dissemination, visualization). Ethics and legal implications permeate each of these. Satisfying capstone projects usually have significant data engineering and data analysis components. Projects might involve some data acquisition, but this is not the primary purpose of a capstone project.

Data Set

Both students and sponsors tend to be dissatisfied with capstone outcomes when there are delays in data access. Plan to have data ready and legal/access issues sorted by the data delivery deadline during onboarding, which takes place in the weeks following kickoff.

The most successful projects involve data sets that are well-defined, so a data dictionary is a great help.

Sponsor Engagement

The most successful projects have strong advocates within the sponsor organization. This sponsor liaison is responsible for communicating with the student team, providing context, direction, and history, and mobilizing other parts of the organization to provide information and data to the team. If the sponsor liaison is not familiar with the data (especially the access, security, structure, and definitions), it is helpful to identify a second liaison who can answer technical questions.

Multiple Points of Contact

Many capstone projects move forward smoothly with a single primary point of contact. In some cases, however, the main sponsor may benefit from support from others within the organization to help enable the project. For example, the person defining the research question may not be the same individual who manages or grants access to the data. When a data owner is involved, their role is typically most important early in the project to support data access and initial setup. For projects that bring together data from more than one organization, there may be a data contact at each partner organization.

Occasionally, projects also involve coordination with contracts or legal teams, particularly when formal agreements or invoicing are required. These stakeholders are generally involved only during the setup phase, and we work to keep engagement as lightweight as possible. 

Time Commitment Close Icon Close

Generally, sponsors are expected to commit one hour per week to meeting with students/replying to student correspondence. There is also a virtual project open house meeting with students in November or April, depending on the project cycle. Finally, sponsors are encouraged to attend and participate in the capstone showcase in May or August, depending on project cycle, where students present their work. 

Eligible Organizations Close Icon Close

Any organization, whether corporate, nonprofit, academic, governmental or otherwise, is eligible to sponsor a capstone project.

Sponsor Open House Close Icon Close

The Sponsor Open House gives students an opportunity to learn about proposed capstone projects and engage directly with prospective sponsors. Each sponsor is asked to prepare a short slide presentation covering organizational background, the problem to be addressed, and anticipated project deliverables, followed by time for questions and discussion.

Each sponsor should plan for a total presentation time of approximately 15 minutes, including Q&A. Projects presented at the Open House are typically staffed following the event.

Intellectual Property Close Icon Close

UVA policies state that students own the IP for projects completed in courses.  The School of Data Science encourages the distribution of such work under appropriate open source (software) and/or Creative Commons  license(s). However, many sponsors prefer that the products of the capstone work remain the intellectual property of the sponsor. In these cases, sponsors ask the students to sign an IP agreement and an NDA.

Confidentiality and Nondisclosure Close Icon Close

Students are instructed on privacy, security, and ethical uses of data and analysis. UVA is equipped with computing infrastructure to work with sensitive and protected data on site if necessary. If students will be working with personally identifiable information (PII) or personal health information (PHI) as a part of their project, we recommend connecting with the capstone director before the start of capstone.

If the sponsor is willing, students may publish and present results. Other agreements related to data use and intellectual property can usually be worked out.  Contact Adam Tashman (DataCapstones@virginia.edu) with any questions about such agreements and addenda. 

Example Capstone Projects

View a sample of recent capstone projects.