Robotic Process Automation – Underwriting in Life Insurance for New Business – 1

Underwriting in Life Insurance for New Business
Recently, we are seeing a lot of interest being generated around Artificial Intelligence, Machine learning and its specialized subset “Deep Learning” or rather Neural Networks to solve a wide variety of problems. The application of machine learning now covers a wide spectrum of industry domains. While the concepts of neural networks have been there for a while the development in computer hardware and processing power, Internet of Things (IoT), abundant availability of data have made possible their implementation and applicability across industries. Read More

Robotic Process Automation – Underwriting in Life Insurance for New Business – 2

Underwriting in Life Insurance for New Business

This blog is in continuation to my previous blog on Introduction to Robotic Process Automation for Underwriting in Insurance New Business.

People familiar with the analytics space would have heard of affinity analysis or association analysis. These analyses are very frequently used in the retail domain particularly in analyzing the propensity of a customer in buying an item given he has already brought another item(s). Such an analysis in the retail domain is known as a Market Basket analysis and has provided some amazing insights on the buying pattern. Perhaps, one of the most interesting stories related to the affinity analysis is the sale of beer and diaper at Walmart which has intrigued many analysts. Affinity analysis has found applications not only in retail but many other domains as well. Read More

Custom Predictive Modeling Dashboard Using Spotfire

Analytics based techniques enables us to improve decision making strategies in real time with help of using machine leaning models generated from historical data and analyzing their outcome on real unclassified data

CAPIOT analytics team have developed an indicative Spotfire Dashboard which can be used to generate, compare and deploy the predictive models. The following are the key components of the dashboard

  • Perform descriptive analysis of the data
  • Configure and generate machine learning models
  • Compare and deploy the most optimal model for generating risk score at real time.

The model is deployed on TIBCO Streambase for real time scoring on streaming data. Read More

Integrating BWPM Data in Spotfire and KIBANA

Many of us that used older technologies for reporting like Excel sheets are now accustomed to new interactive environment ma for reporting – tools like Spotfire.

As tools tend to look more interactive, people’s imagination and expectations grow. Their reaction to these new age reports are typically: “This is okay, but I want some thing more descriptive and easily understandable”, in spite of having managed for years and years on excel. Read More

Runtime HashMap Refresh for TIBCO BW

Even though the ESB is purely for Integrating two web services. Sometimes, ESB would perform Data Massaging and transformation in case if the Consumers and providers of the web services couldn’t synch.

ESB transformation of data requires a Lookup from external storage. But since each process has to wait for certain clock cycles to fetch data from External storage, we used an approach of loading Lookup data into Cache for a quick retrieve. Whenever the BW Component is restarted the lookup data will be loaded into cache. But still, this approach has below disadvantages, Read More

GodMode logging

The following post explains, how to design a BW process that receives and processes the request from UI to toggle the payload mode. The below design eliminates the usage of RV palette / RV command execution which eases the manual work and also it is easy to couple with UI. Also, this design allows processing the request in batch.

God Mode

The process of enabling or disabling the payload in ESB by triggering an RV message on a subject is what we call GodMode.  This toggles the payload mode smooth by without bringing down or restarting any components. Read More

Importing ElasticSearch Data into TIBCO Spotfire using R

ElasticSearch is fast-becoming one of the most widely adopted document storage solutions for machine data, especially when combined with tools like logstash that allow easy collection of information from log files. However, while Kibana (another open source project part of the ElasticSearch kit) is a very popular data visualization tool, TIBCO Spotfire remains one of the top tools in its ability to slice and dice data.

There are several situations where importing Elastic Search data into TIBCO Spotfire will yield interesting results. This blog post aims at covering one such example which explains how to bring data from Elastic Search into TIBCO Spotfire. Read More