University of Chicago · Data Science Institute
Research software engineering, data science, and AI for the public interest.
We partner with researchers, fellows, and mission-driven organizations to turn prototypes into robust, reproducible, real-world software.
- Projects
- 13
- Partner organizations
- 14
- Engineers, scientists & admins
- 13
- Talks & posts
- 09
01 — What we do
From research prototype to production software.
01
Research software engineering
We partner with researchers to design, develop, and maintain robust software tailored to academic research, applying best practices to turn prototypes into production-ready tools that are efficient, reproducible, and usable by the broader scientific community.
02
Data science & AI
Our data engineers, data scientists, and system administrators work closely with faculty and research teams to design workflows, build data pipelines, and operationalize advanced AI models on cloud infrastructure and the DSI computing cluster.
03
Tool building
We build interactive software, web applications, and AI tools that make research insights accessible and actionable: dashboards, AI agents, public-facing platforms, and domain-specific applications that amplify the impact of your work.
02 — Featured projects
Work with real-world impact.
03 — Our mission
Better code, broader impact.
We exist to make the computational side of research faster, cheaper, more reproducible, and more useful to the people it's meant to serve.
- 01
Expand the impact of DSI research through high-quality code, lowering the barrier for academic, government, and industry researchers to adopt new tools.
- 02
Support the translation of DSI projects into practical applications.
- 03
Strengthen grant proposals by increasing the resources available to faculty.
- 04
Decrease the cost of the computational side of data science research through a centralized hub for technical resources.
- 05
Deliver social impact through collaboration with nonprofit and community organizations.
- 06
Educate a diverse and inclusive next generation of ethically minded data science leaders.
04 — Lunch & Learn
What we're learning.
Slides and notebooks from our internal talks on AI foundations, performance, cloud, and tooling.
All talks-
The Transformer
Jim Pivarski
This is part of our lunch study on the foundations of LLMs. This talk explains all the parts of the transformer architecture and why each is needed.
-
Backpropagation
Jim Pivarski
This is part of our lunch study on the foundations of LLMs. This talk explains how derivatives for optimization can be computed efficiently, starting from simple numerical and symbolic approaches and building up to backpropagation.
-
Agentic Coding Tools
Nick Ross
A map of the agentic coding tool landscape: how the tools differ, how they are used, and an opinionated take on how the team does or does not use them.
05 — Work with us
Have a research problem that needs software?
Tell us about the research question, the data, and what "done" looks like. We'll help scope it.