A national security services provider engaged Barnes & Thornburg to help assess and limit possible liability in connection with a potential class action involving more than 2,000 putative class members. The matter team included Partner Mark Wallin, Associate Ben Perry, Data Operations Engineer/Analyst Janina Spencer, and Client Solutions Coordinator Kim Batten.
Challenge
The team needed to pull licensure data from a public-facing government database, compare it to relevant employment and damages periods, preserve supporting evidence for each result, and protect sensitive class-member information. Because the database required individual searches and did not offer a bulk export, a manual review was not practical at this volume or within the project timeline.
Approach
After determining that available AI platforms were not the right fit because the task required browser automation and data extraction from an external website, Barnes & Thornburg assembled a project team of attorneys and automation specialists to develop a custom, locally run browser automation tool.
The Python-based tool allowed the team to search for licensure status and history for everyone, and save a screenshot of each result. The workflow flagged unsuccessful or ambiguous searches for manual review rather than accepting them automatically. It also generated structured data files, preserved source evidence, and compiled normalized results into an Excel workbook.
The results were then matched to class-member records and applied to relevant exposure periods to support the damages analysis. The process ran locally on a firm-managed computer, and the final output was delivered through an approved firm storage location, allowing for the preservation of security and confidentiality.
Impact
The automated workflow completed in about 12 hours of unattended runtime. A manual review was estimated at six to ten minutes per person — roughly 200 to 330 hours for more than 2,000 individuals. The process eliminated repetitive searching and data entry while preserving source evidence for quality control. That allowed the team to focus on exceptions, review, and the substantive exposure analysis.
