Consensus Corporation simplifies the complex process of selling connected devices, such as cell phones, by unifying retailers, manufacturers, and network operators under a single platform. One of its core services is utilizing machine learning for faster and more accurate fraud prevention, or alerting retailers to high-risk customers before they purchase expensive devices. To identify potential fraud, Consensus built an advanced data model for machine learning that leverages huge volumes of data and undergoes routine updates. Consensus sought out a more efficient means to prepare data for its model to increase the accuracy and immediacy of its fraud detection.
- Preparing data for machine learning was time-consuming. Due to the painstaking process of reengineering SQL scripts, it took Consensus three to four weeks to update its fraud model.
- Data inaccuracies risked millions in fraudulent purchases. With each percentage lost in fraud detection, retailers lose roughly a million dollars in profit
- Consensus product and BI teams weren’t able to prepare data themselves. Without advanced SQL knowledge, preparing data required multiple requests to IT.
Solution with Trifacta
- By deploying Trifacta Wrangler Pro through the Amazon Web Services marketplace, Consensus utilizes Trifacta’s close integration with machine learning platform DataRobot to reduce data wrangling and model development time from four weeks to mere days.
- With the ability to more quickly improve its fraud detection model, Consensus have saved retailers millions in fraud loss
- The intuitive experience of wrangling data in Trifacta has allowed Consensus product, machine learning and BI teams to work directly with raw data, limiting IT requests
Consensus Corporation is radically simplifying the decidedly un-simple transactional process of selling connected devices, from point of sale through activation and beyond. Consensus is a subsidiary of Target Corporation and headquartered in San Francisco.