The Data Lab partners with organizations across child welfare, juvenile justice, and adjacent fields to answer the questions these systems face. Privacy is built into how we work: our methods and infrastructure are designed to safeguard sensitive information, ensuring the highest level of confidentiality and security.
Get in touchAPPROACH
We use AI, but we do it differently. Every analysis runs on our own hardware, inside the lab. Sensitive records never leave our walls, never go to outside companies, and never touch the cloud. That is the only responsible way to handle confidential data about children and families. Working at this scale also uses far less energy than the massive AI systems in commercial data centers; a smaller footprint, by design.
From the right questions to answers you can act on.
We help organizations write surveys that actually get useful answers, and then make sense of what comes back. We work on questionnaires for program evaluations, staff and stakeholder feedback, and research studies; the goal is to ask the right things in the right order so the data you collect is data you can act on.
A steady way to gather information over time.
We build the systems that gather your information consistently, year after year. That includes intake forms on the web, surveys that work on a phone, ways to pull information out of existing administrative records, and methods for connecting records across different systems. The point is something the people on the ground can actually keep using.
Working tools for working professionals.
When an organization needs a tool that does not exist yet, we build it. A recent example: a real-time bed-tracking application used by state staff to find juvenile justice placement availability across the system. We design tools that fit the way people already work, instead of asking practitioners to change how they work.
Making patterns visible to the people who can act on them.
We turn complicated information into dashboards and charts that the right people can read at a glance: a court administrator, a program director, a policymaker. The goal is always the same; make the picture clear enough that someone can actually do something about it.
Reading thousands of records when reading one at a time isn't possible.
Many of the most important records in social services are written by hand: case files, investigation summaries, court notes. We use AI to read through tens of thousands of these documents at once and surface patterns no person could find on their own. The Data Lab has published peer-reviewed studies using this approach to identify substance use, domestic violence, firearms exposure, and other concerns inside child welfare records.
Studies that tell you whether something is actually working.
We help organizations figure out whether their programs and policies are doing what they were meant to do. That means choosing the right outcomes to look at, picking comparison groups carefully, and being honest about what the data can and cannot tell you. The result is evidence that holds up, and that answers the question you were really asking.
GET IN TOUCH
Whether you have a defined project in mind or a question you don’t yet know how to ask, the lab welcomes an early conversation. We work with state and county agencies, courts, non-profits, and research partners both inside and outside the University of Michigan.
ssw-datalab@umich.edu