
Zach DeBruine
CEO, product and technology
Zach holds a Ph.D. in machine learning and serves on the AI faculty at Grand Valley State University.
zach@contigtech.comNow deploying with Michigan law enforcement agencies
Contig answers questions from your agency's reports, case files and records requests. It runs inside your CJIS boundary. Every answer shows the page it came from. People make every decision.

Swipe sideways to see the cited source.
Each sentence links to the record it came from.
One click opens the passage with the cited text highlighted.
Contig points out what the records leave unanswered.
Each question and each source you open goes into your agency's audit log.
The problem
“Right now, it's just purely on memory, and it takes a lot of time to get to that.”
17,541
Bureau of Justice Statistics, Census of State and Local Law Enforcement Agencies, 2018
1 in 4
Bureau of Justice Statistics, Multistate Criminal History Patterns of Prisoners Released in 30 States, 2015
20%
Birkenfeld, Homeland Security Affairs, 2021
Contig gives every person in the agency the memory of the whole agency. Every answer traces back to its record.
What it does

Our standards
We monitor KPIs with you and your team. We use your feedback to optimize our AI systems in real time.
Each citation supports the sentence beside it.
Contig answers from your records and flags questions outside them.
Proposed redactions find personal identifiers and exempt items.
Each address lands where the record places it.
Answers arrive fast enough to use in the middle of a case.
Security and principles
Each agency gets its own Azure Government environment. People sign in through your identity provider. Every question, source and decision goes into an append-only log. Your agency owns that log and can export it. Your records stay out of model training.
Contig works on records your agency already holds.
The software proposes. Your staff decide under their own names.
Every action is recorded and ready to export at any time.
Export it whenever you like. We certify deletion when you leave.
Email us for the security documentation your CJIS security officer needs →
The software proposes. A person decides. Everything is logged.
Next: the first connection between two Michigan agencies
Neighboring agencies that both run Contig agree on a search policy and sign it. Contig flags possible connections across their records. An investigator reviews each one. An officer approves each request on a single case. Both sides log every step.
How an agency starts
We meet the people who would use Contig and the team that owns its environment.
We sit with detectives and records staff. We watch the work before we propose anything.
We define the work together. Both sides agree on what finished means.
We build. You review and accept each capability before the next one begins.
Who we are
Contig began when a Michigan sheriff's office brought a problem to Grand Valley State University. Its detectives were connecting cases from memory. We built the answer with them.
We start with law enforcement in Michigan. Every public agency that keeps records and answers to the public faces the same problem.

CEO, product and technology
Zach holds a Ph.D. in machine learning and serves on the AI faculty at Grand Valley State University.
zach@contigtech.com
COO, operations, strategy and partnerships
Aziz founded Innorov and led it to an exit. Aziz is earning an M.S. in applied computer science at Grand Valley State University.
aziz@contigtech.comEmail us to book a briefing, request our security documentation or talk about working together.
Email us