Help reentry attorneys and paralegals pull a client's complete criminal court records from across counties, structure the data, flag warrants and risk items, and produce a cost-and-fee analysis report that supports case strategy.
This page is the shared target for this situation. Any court, legal aid office, or builder working on it in their own region can take the vision, the blueprints, and the standards here instead of starting over, and set the parts that are local to them. The point is that the same work is not designed from scratch in every county.
Does this match how it goes where you are? Leave a note on this workflow.
Here is where each shared asset stands for Criminal Records Data Assembly and Analysis. The reference architecture and the evaluation protocol are the furthest along, drawn from an active implementation with a legal aid partner. The remaining assets will be drafted as the cohort forms.
What a criminal records assembly system needs to do, from searching court databases and resolving a client's identity across counties through extracting case data, classifying warrants and fees, and producing a structured report. Written so a team can scope a build, evaluate a vendor, or write requirements for a grant proposal.
Coming soonThe quality bar a system must clear before it should be trusted with real client records. Covers whether it finds all cases for the right person, whether it correctly classifies warrant types and risk levels, whether financial totals are accurate, and whether it flags what it could not retrieve rather than guessing.
Coming soonA structured review process for scoring a records assembly report against the actual court data. The protocol walks a reviewer through emergency flag checks, case landscape review, probation and deferment verification, and source-of-truth comparison for cases, warrants, and payment amounts. Includes a guided review form and auto-fail conditions for safety-critical errors.
Join the cohort to contribute →How to build a system that searches court records databases, extracts structured facts from HTML docket pages, analyzes fees and warrants per case, and produces a report an attorney can use. Describes a two-phase pipeline: a lightweight model parses raw court pages into structured data, then a heavier model reasons over those facts to produce case-level analysis and cross-county summaries.
Join the cohort to contribute →The split between the parts any jurisdiction can take as they are, such as the structure, the checks, and the evaluation, and the parts each place has to set for itself, such as the statutes, the forms, the deadlines, and local practice. A team should be able to see which is which before it starts building, so it adapts what exists rather than rebuilding it.
Join the cohort to contribute →Write-ups from teams who have already built something here: how they put it together, what they would do again, what they would not, and the materials they can share. If your team has one of these running, or half built, tell us about it and we will write it up for the next team.
Join the cohort to contribute →The Court Wrangler, built by a university research team with a legal aid partner, assembles criminal court records from the statewide court record systems for reentry clients facing outstanding fines, fees, and warrants. The case study covers what the team chose, how the tool was tested, what the attorneys needed from the reports, and what changed across ten months of biweekly iteration.
Join the cohort to contribute →Evaluation scores from running the test suite against the Court Wrangler and any future implementations, published on JusticeBench. Results will include rubric scores by category, error rates by type, and verification time per report.
Coming after test suite is runA guided review application that walks a QA reviewer through each step of scoring a records report, with inline reference guides for warrant classification, probation terminology, and payment verification. Also includes an Airtable-based scoring form for collecting structured evaluation data across reviewers and testing rounds.
Join the cohort to contribute →A course that walks a team through building, testing, and deploying a criminal records assembly tool for their jurisdiction and court systems, using the assets on this page as the course materials.
Coming as assets matureThe resources build on each other. The functional agenda defines what the system must do. The conformance standards define how well. The test suite provides the cases to verify it. The reference architecture shows one way to build a system that meets the specification. The case study shows how one team did it in Oklahoma.
You do not need all of them. Pick the resources that match your role.
A person with a criminal record in Oklahoma may have cases spread across dozens of counties, each with its own docket, its own fee structure, and its own warrant status. A reentry attorney preparing to help that person needs to know the full picture: every case, every outstanding balance, every active warrant, every probation status. Today that means hours of manual searching across court websites, clicking through case after case, copying numbers into a spreadsheet, and cross-referencing payment records.
Criminal records data assembly is a system that searches court records databases for a given client, pulls all their cases across all counties, extracts structured data from each case (charges, dispositions, fees, payments, warrants, probation), and produces a report the attorney can use to make decisions. The system handles the data gathering and structuring. The attorney handles the strategy, the judgment, and the client relationship.
The workflow applies to any jurisdiction where criminal court records are published online and where attorneys, paralegals, or legal aid staff spend significant time manually pulling and organizing that data. The Court Wrangler, built with a legal aid partner, is the first implementation. It pulls from the two statewide court record systems and produces a Word document report organized by county, with an executive summary, warrant flags, payment history, and a list of follow-up items for the attorney.
The analysis layer goes beyond assembly. The system classifies warrants by risk level (distinguishing an Order for Arrest from a Cite and Release), identifies probation and deferred sentence status, computes per-case and cross-county financial totals, and flags items that need manual follow-up (such as DA fees, supervision fees, and pre-2005 records that are not available online). The attorney uses the report to answer three questions: Can we bring this person before a judge safely? Can we represent them or do they need a public defender? What is the strategy for reducing their debt?
Who this serves: Reentry attorneys, paralegals, and law students at legal aid organizations, public defender offices, and pro bono clinics that help people with criminal records address outstanding fines, fees, warrants, and eligibility for expungement or record sealing.
The Criminal Records Data Assembly working group is developing the shared assets on this page. A university research team is leading the implementation work with a legal aid partner, and the cohort will shape the standards and resources so other jurisdictions can build on what Oklahoma learns.