Case Study: Consensus Corporation achieves 24% improved fraud detection and 55% fewer false positives with DataRobot

How Consensus, a Target subsidiary, simplified data wrangling for machine learning

Consensus Corporation, a Target subsidiary that helps retailers sell connected devices, faced slow, resource‑intensive data wrangling for its fraud detection models. Preparing and re‑engineering large, disparate datasets took weeks and required advanced SQL and data‑science skills, limiting the company’s ability to iterate quickly. After evaluating options, Consensus selected DataRobot’s automated machine learning platform together with Trifacta Wrangler Pro on AWS to simplify data preparation and enable faster model prototyping.

Using Trifacta to wrangle historical data in Amazon S3 and DataRobot to automate model building, validation, tuning, and deployment, Consensus cut feature engineering and model deployment from weeks to hours. The solution delivered measurable impact: a 24% improvement in true positive fraud detection, a 55% drop in false positives, a 19% gain in overall financial performance, and model training times reduced from days with multiple people to 2–3 hours by one person. DataRobot’s platform was central to achieving these gains.

Consensus Corporation

Harrison Lynch

Senior Director of Product Development