Winter 2026 Capstone Showcase Highlights Innovative Data Science Projects

The Winter 2026 Capstone Showcase for the University of Chicago’s Master’s in Applied Data Science (MS-ADS) program featured 14 student teams presenting projects that applied data science to real-world challenges.

Across two presentation sessions, students presented solutions spanning industries such as healthcare, finance, supply chain, marketing analytics, robotics, and environmental policy. Capstone projects represent the culminating experience of the MS-ADS program, where students design and implement end-to-end data science solutions.

Following the presentations, judges selected Best in Show winners from each session, along with honorable mentions recognizing additional outstanding projects.

Best in Show Spotlights

Here is a closer look at the top projects recognized by the judges. 

Dinkster: Computer Vision Pickleball Coaching

Presenters: Mo Abdelhamid, Victor Doddy, Kevin O’Leary, Pat Ohea


Faculty Advisor: Nick Kadochnikov


Pickleball has evolved from a niche retirement community pastime to a national obsession, yet the infrastructure for mastery hasn’t kept pace with its popularity. For the casual player, professional coaching remains a luxury of both time and deep pockets. A new capstone project, Dinkster, aims to democratize the “pro” experience by turning raw smartphone footage into a sophisticated biomechanical post-game analysis.

The interface is simple: a player uploads a video of their match, and a computer vision pipeline goes to work. Behind the scenes, the system dissects the chaos of a rally, tracking the arc of the ball and the footwork of the players, and filters that data through a Large Language Model to deliver actionable coaching cues.

By benchmarking amateur movement against professional standards, Dinkster identifies the technical inconsistencies in a player’s game. It analyzes everything from powerful baseline drives to the delicate, strategic shots at the net that define high-level play. As faculty advisor Nick Kadochnikov noted, the team’s success lay in navigating the “real-world complexity” of the sport. What impressed Kadochnikov the most was how thoughtfully they handled real-world challenges like separating players from spectators, tracking a specific player on a crowded court, and adapting to different camera angles.

The system also analyzes player poses across different phases of play, from baseline drives and third shot drops to kitchen dinks and volleys, and converts those biomechanical signals into targeted recommendations that help players refine technique and improve performance. While the project concludes with this capstone, it points toward a future where elite sports science is available to anyone with a camera and a court.

Data for the Common Good: EPA Capstone

Presenters: Lauren Adolphe, Brianna Ngo, Rinad Salkham, Aneesha Dasari


Faculty Advisor: Nick Kadochnikov

The second Best in Show project focused on environmental accountability, tackling the challenge of independently verifying industrial emissions, as regulatory oversight can shift over time.

The team developed a system that integrates facility data with observations from the European Space Agency’s Copernicus Program, a sophisticated Earth observation mission. Central to this approach is a data fusion pipeline combining ground-based sensor inputs with imagery from the Sentinel-2 satellites, which provide high-resolution optical data used for everything from agricultural monitoring to water quality, and Sentinel-5P atmospheric data, which tracks trace gases in the air.

By aligning these multiple data streams, the system can estimate facility-level nitrogen dioxide (NO2​) concentrations and flag anomalies where ground-level reporting doesn’t align with what the satellites see from orbit. This introduces a more comprehensive view of emissions patterns and potential underreporting.

According to faculty advisor Nick Kadochnikov, the project is especially relevant in a shifting regulatory landscape. By leveraging globally available satellite data that operates independently of U.S. policy changes, the system introduces a more resilient and transparent method for monitoring emissions.

Beyond its technical implementation, the project shows how data science can be applied to public-interest challenges, surfacing environmental risks, improving accountability, and bringing greater visibility to the communities most impacted by pollution.

Honorable Mentions

Here are additional projects recognized by the judges that demonstrated strong innovation and real-world impact. 

Fine-Tuned LLaMA Agents for Autonomous Quantitative Investment Research


Presenters: Alejandro Canete Baez, Sebastian Rivera, Sami Naeem, Campbell Taylor


Faculty Advisor: Justin Kurland

SmartAudience360: A Data-Driven Platform for Precision Targeting & Conversion Lift

Presenters: Prinu Mathew, Qingwei Zhang, Joshua Carbajal


Faculty Advisor: Nick Kadochnikov