ENVISION Predictive Modeling

Knowledge Examples
Loss Reserving
Why Us?

Envision Enterprise Predictive Modeling and Service

ENVISION is an enterprise level predictive modeling  platform designed for sophisticated calculation and big data analysis. It is powered by open source R, SQL Server, parallel computing, and grid scheduling techniques.


 We can help you analyze your data and setup analytical models

 We customize Envision to fit your business process

 Or you can simply use Envision to empower your own analysis

Supervised Learning models 

Unsupervised Learning models 

Big Data - Grid Scheduling and Pull Technology

ENVISION innovative Grid Job Scheduling and Pulling Scheme allows users to make runs at the same time using multiple machines for multiple models. A pool of jobs is created by the user and available computing servers pull jobs from the pool until it is empty. The results are then summarized, compared, and presented.

Big Data - Random Ensemble on GLM, ANN and CART

Random ensemble for GLM, ANN and CART models, constructed upon the Parallel Grid Scheduling technique, is another innovative solution to efficiently overcome the big data challenge. The model slices the big data into small random datasets, then each subset is used to train a supervised learning model. The final prediction results are based on the average results of all trained models.

Successful Case Studies

  • Sinolife Ins. Case Study 
  • Cancer Mortality Prediction 
  • Usage Based Insurance (UBI) 

Envision Insurance Applications

Pricing Reserving Underwriting Marketing Risk Analysis Business Disruption
More pricing factors
Claim classification
Case reserve adequacy assessment
False claim identification
Claim closure/reopen
Claim size prediction
High risk case identification
Automatic underwriting
Customer retention, renew, or resell
Personalized product recommendation
Customer categorization
Fraud detection
Credit score
Health/Fitness discount