Submit a project to the X-Lab
BU Spark! offers a unique opportunity for individuals and organizations to partner with Boston University students with computing, data science, and engineering skills on their technology projects. This includes, but is not limited to, analyzing data, testing machine learning models, and developing apps. There are several no-cost or low-cost options for partners to engage with students. Read our FAQs for more information.Submit a project
Free class projects
Justice Media co-Lab: The Justice Media co-Lab (XC410 A1), is a collaboration between Spark! and the BU Department of Journalism that matches interdisciplinary student teams with backgrounds in computing, data science, statistics, engineering, journalism, and related disciplines with computational journalism projects. Projects vary in size and scope, and range from 60 hours of work to 100 per team member over the course of a semester. Each project is usually assigned three computer science or engineering students and one to two journalism students. Partners usually dedicate around 45 minutes per week for the eight to 12 weeks of project implementation.
Data science classes: Data science projects help partners answer strategic questions from data analysis conducted on multiple large data sets, usually comprised of a minimum of 5,000 records. Projects must include data collection (such as calls to an API, parsing or crawling web pages), data complication and cleaning (such as combining with an internal spreadsheet), and analysis. Students analyze data using methods like clustering, classification, regression, and network analysis. Partners should have a specific question(s) that they want answered from analysis. Example: In partnership with the ACLU of Massachusetts, Spark! students identified patterns of racial and geographic disparities of the treatment of residents by the Boston Policy Departments. Students used stop and frisk, crime incident report, and census data.
- Data Science Tools and Applications (CS506)
- This course is best for projects that look to analyze data that is updated annually. Each year students in CS506 will repeat projects from previous years using updated data and questions from stakeholders to provide new insights. Each project may be assigned multiple teams of four to five students.
- Tools for Data Science (DS701)
- This course is best for exploratory data science projects. Teams of four will explore the data provide and use biweekly meetings with partners to focus their analysis and provide relevant insights.
- Data Science for Good (XC410 B1)
- This course matches interdisciplinary student teams with backgrounds in computing, data science, statistics, sociology, economics, psychology, and other related disciplines with data science projects focused on the criminal justice system. Projects for this course must have a connection to the justice or criminal legal system in addition to meeting the data science project criteria above. Teams of four to five will meet with partners biweekly to share insights and get at the questions partners presented.
Data Visualization class: Students in Spark!’s Data Visualization practicum (DS549) apply data visualization practices to data and questions provided by partners. Partners must already have compiled and cleaned data set comprised of a minimum of 5,000 records and have a specific question(s) that they want answered or displayed through the analysis and visualization. Teams of three or four will meet biweekly with clients to explore these questions and review the work done by student teams.
Machine Learning class: Machine learning projects build algorithmic models that accurately forecast desired outcomes through a supervised learning approach. Partners must already have compiled and cleaned data set comprised of a minimum of 5,000 records, including a portion of “ground truth” data, that is, where the result is known. Students in Spark!’s Machine Learning practicum (CS/DS549) use ground truth data to build a model that trains the remaining data to achieve forecasts based on desired thresholds of accuracy determined by the partner. Example: Using data pulled from the Zillow API and RMLS data, students created an algorithm to estimate real estate sales prices for an early-stage startup. Variables included interior and exterior photos, location, and house features. The model built achieved a 81% accuracy rate.
Software Engineering class: Students in Spark!’s Software Engineering practicum (CS/DS519) work on projects that have a focus on software engineering. Partners come to Spark! with an idea that is ready for development, and our team will work with you to determine feasibility and the best approach. Projects in this course range from mobile app development to frontend and full stack development. Teams of three or four will meet biweekly with clients to develop the product partners are looking for.
User Experience (UX) Design class: Students in Spark!’s UX Design practicum (DS488) engage in projects at different stages of UX research and design. Partners come to Spark! with a product they want to design or an idea they want to explore starting with user interviews, and our team will turn that into a project plan for students. These projects can include the development of UX prototype and sketch wireframes, style and brand development, and if desired, Hi-Fidelity wireframes. Teams of three or four will meet biweekly with clients to develop the product partners are looking for.
Paid, on-campus externships projects
Partners can fund a student team through an on-campus externship. These projects offer greater flexibility both in terms of technical scope and timing as these projects may extend beyond the semester timeline. Submit a project or email us to learn more.
Read our FAQs