Customer
Large agroholding with 1,5+M acres of own land, one of the market leaders in in sugar, pork, crop, oil and fats production.
Challenge
Design and develop the real-time data analysis platform that detects fraud and anomaly cases during harvesting operations. The specific algorithms applies on the streaming data from trackers installed on agricultural machinery and the data from harvesting process control systems and ERP. As a result, the platform identifies abnormal behavior and sends alert with all required data.
Project Results
The client receives a possibility to access in real-time to comprehensive set of harvesting operational data and apply different ML models and algorithms to identify fraudulent activities and deviations from the standard procedure.
Business Value
Reduced losses due to fraud and improved productivity.
Tech Stack
Apache Hadoop, Apache Spark (Streaming, MLlib), Apache HBase, Apache Zeppelin, Apache Kafka, Akka.
Large agroholding with 1,5+M acres of own land, one of the market leaders in in sugar, pork, crop, oil and fats production.
Challenge
Design and develop the real-time data analysis platform that detects fraud and anomaly cases during harvesting operations. The specific algorithms applies on the streaming data from trackers installed on agricultural machinery and the data from harvesting process control systems and ERP. As a result, the platform identifies abnormal behavior and sends alert with all required data.
Project Results
The client receives a possibility to access in real-time to comprehensive set of harvesting operational data and apply different ML models and algorithms to identify fraudulent activities and deviations from the standard procedure.
Business Value
Reduced losses due to fraud and improved productivity.
Tech Stack
Apache Hadoop, Apache Spark (Streaming, MLlib), Apache HBase, Apache Zeppelin, Apache Kafka, Akka.