The Data Science for Social Good Fellowship is a full-time summer training program run by Carnegie Mellon University and the Data Science for Social Good Foundation, hosted each year on a US campus. The 2026 edition ran in collaboration with and at Johns Hopkins University in Baltimore, Maryland. Fellows work in teams on machine learning, data science and AI projects scoped in advance with government and nonprofit partners, in areas such as urban blight, fire safety, building vacancy and traffic policy.
Each team is supported by full-time senior data science mentors and project managers, and project work is combined with lectures, hands-on workshops and seminars. The program describes its aim as producing practitioners who can solve real-world problems fairly and equitably, and past cohorts have worked across criminal justice, health, education, economic development, infrastructure and social services.
Benefits & funding
- Paid fellowship: the program states stipends are fixed and vary with the cost of living of the host location, and should be more than enough to cover travel, housing and other expenses for the summer. No dollar figure is published.
- On-campus housing arranged by the host university (2026: Johns Hopkins, move-in 24 May 2026).
- Full-time senior data science mentors and project managers assigned to each project team.
- Lectures, hands-on workshops and seminars alongside the project work.
- Real projects built with government and nonprofit partners; many teams end up writing papers, though the program says publication is not the goal.
Eligibility
- The 2026 cycle was open only to graduate students currently enrolled at a US university.
- No citizenship requirement, but for 2026 applicants had to already be in the US and require no visa sponsorship — the program could not support visas.
- Graduate students are preferred. Recent graduates from the past couple of years, postdocs and early-career people between degrees are considered; undergraduates only where they show both the computational skills and the drive to keep up.
- Prior experience analysing data with a programming language such as Python is expected. Backgrounds outside computer science are explicitly welcome — teams are built to hold a broad set of skills.
- Full-time and on-site for the whole program; no remote option, with exceptions of up to two days off for critical cases.
Can you apply?
Here's who can apply
Decoded from the official rules. A free account checks them against your passport, degree and age in seconds.
Residence required
USA
Education required
Master Enrolled · PhD Enrolled
Circlebreaker is a free database of global fellowships, scholarships and programs — every one decoded like this.
We only check the criteria we have data for. Always confirm full eligibility on the official page before applying.
Application process
- 1Watch the program site for the cycle’s application page and the dates of its online information sessions.
- 2Submit the online application form linked from the site, describing your skills, relevant projects and your motivation for social-impact work.
- 3Arrange two letters of recommendation from the people who know you and your work best, able to speak to academic background, data science skills, communication and commitment to social impact.
- 4Applications are reviewed on a rolling basis after the deadline; interview invitations typically go out within one to two weeks of close.
- 5Final decisions arrived by mid-March in the 2026 cycle. Selected fellows submit project preferences, and teams are assigned from those preferences and partner needs.
Program timeline
Application deadline for the 2026 cycle (extended from 1 March)
Mar 6, 2026
Final decisions
Mid-March 2026
Housing move-in at Johns Hopkins University
May 24, 2026
Program starts
May 25, 2026
Program ends
Jul 31, 2026