Master’s in Applied Data Science Summer 2026 Capstone Winners
The University of Chicago’s MS in Applied Data Science program celebrated the culminating work of its graduating students at the Summer 2026 Capstone Showcase on August 15th. Across two sessions, student teams presented 14 projects spanning healthcare, finance, insurance, and more – even a home design.
The Capstone Showcase brings together the skills and knowledge students develop throughout the MS-ADS program with real-world problems and emerging areas of research. This summer’s showcase included industry-sponsored, research-focused, and student-proposed projects, with teams developing solutions beyond the classroom.
Best in Show Winners
Research (Agentic AI) | AI Home Design System
Presenters: Aigul Azamat, Akshat Gupta
Faculty Advisor: Nick Kadochnikov
Many people know what they like when they see it but struggle to describe their ideal interior design style in words. The team developed an AI home design system that allows users to upload a photo of their room and explore professionally renovated spaces with similar characteristics. After selecting an inspiration, the system generates a personalized redesign while preserving structural elements of the original room. Users can then continue refining individual elements through conversation with the system.
Research (AI Engineering) | Physics-Regularized Time-Series Encoders for DCE-MRI Kinetic Analysis
Presenters: Gustavo Martinez, Shreyas Tandulwadikar
Faculty Advisor: Batu Gundogdu
The team tackled a challenge in breast MRI: estimating a patient-specific dye-delivery curve when the necessary view of the aorta is unavailable. They developed a simulated dataset that inserts mathematically accurate vessel signals into real patient scans and used it to train and compare deep learning models for vessel detection. The work produced more than 1,000 labeled simulated exams and established a path toward testing the models on real patient scans.
Honorable Mentions
Independent Thesis | Studio Copilot: Teacher-Grounded Multimodal Feedback for Solo Vocal Practice
Presenters: Sagana Ondande, Kaushik Kannan, Cicily Mathew, Joyce Zhang, Haobo Yang
Faculty Advisor: Batu Gundogdu
Voice students spend much of their time practicing without their instructor present. This student team developed Studio Copilot, an AI practice tool that provides feedback grounded specifically in a student’s own teacher’s recorded lessons. The system draws from the teacher’s language and teaching methods to provide personalized guidance between lessons.
Pyxidr | Dynamic Re-Optimization of Insurance-Backed Securities Using Agentic AI
Presenters: Arthur Acker, Eloi Bernier, Grace Rowan, Jose Tovilla Rivera
Faculty Advisor: Francisco Azeredo
The team developed a dynamic portfolio reoptimization framework for insurance asset managers designed to respond to changing market conditions while accounting for regulatory and investment constraints. Backtesting showed recurring improvements in cumulative net statutory income compared with a static buy-and-hold approach. The solution is delivered through a dashboard and conversational agent that helps users interpret the optimizer’s results without participating in the underlying financial calculations.
For MS in Applied Data Science students, the Capstone Showcase marked the end of the program, but not necessarily the end of the work. The projects took on problems that extend well beyond the classroom, from medical imaging and financial optimization to personalized AI tools, and for some students, the ideas and solutions developed will continue to evolve long after graduation.