Devron is a federated data science platform that enables teams to build and train models on distributed, heterogeneous, and private data where it resides.
In this demo, we'll walk you through the model-building process and show you how Devron can be used to train a model across horizontally split (same schema) datasets without moving or exposing the data.
We will follow an example scenario where we assume the role of a corporate analytics team looking to predict fraudulent events with customer credit card transaction data from three regional banks.
Use Case Summary:
INDUSTRY: | Banking |
OBJECTIVE: | Fraud Detection |
BUSINESS VALUE: | Minimize Risk |
Preserve Revenue |
Unlock access to valuable, previously inaccessible datasets (including securely and privately sharing data between organizations | |
Analyze sensitive data without the risk of privacy leakage or lineage issues | |
Boost model accuracy, generalizability, and reliability | |
Accelerate time to insight by drastically reducing data engineering overhead |