BANKINg AND insurance
verticals
Banking
objective
To parse, clean and enrich data from various sources and convert the data into a standardized format allowing for faster data analysis.
solution
Business Intelligence:
Data parsing and enrichment
solution details
- Prepared the various standards acceptable for all fields coming in from source systems
- Developed automated macros to parse, enrich and update data fields from source systems
Banking
To identify exisiting customers who would have a higher propensity to buy other products.
Sales and Marketing:
-Segmentation
-Propensity Modelling
- Identified the processes which had the most complaints
- Conducted sentiment analysis to identify the priority of complaints
- Developed a predictive model to prioritize complaints
Banking
To identify the key processes where most complaints were raised and also identify the degree of sentiments in order to predict if a customer complaint will be raised.
Sales and Marketing:
-Segmentation
-Predictive Modelling
-Text Mining
- Identified the processes which had the most complaints
- Conducted sentiment analysis to identify the priority of complaints
- Developed a predictive model to prioritize complaints
Banking
To develop a propensity to default model on credit card applications.
Strategy and Operations:
-Segmentation
-Propensity Modelling
- Using credit application and bureau variables a model was built to segment customer applications into various risk levels
Banking
To develop response models to identify campaign success metrics.
Strategy and Operations:
Predictive Modelling
- Segmented the historical customer data using RFM methodology
- Performed variable and feature selection to optimize model accuracy
- Developed logistic regression models to identify customers who should be targetted for the campaign
Banking
To develop a call volume forecasting model to manage call centre operations.
Strategy and Operations:
Forecasting Models
- Gathered historical data including factors affecting business variability
- Decomposed the data to identify trends and seasonality in the data
- Developed multiple models using holt-winters and SARIMA techniques
Banking
Identify factors that influence high attrition and poor performancefrom among the current data/information collected at the time of hiring. Create a filteration matrix with the identified matrix which can then be used for future recruitment drives.
Human Resources:
Workforce Modelling
- Using proprietory Redwood algorithms 4 variables were identifed which had the highest impact on attrition
- Employee referral source was identified as a variable of high importance as it negatively impacted performance/attrition. Thisinsight was garnered as a result of very detailed data mining
Insurance
Estimate market size and opportunity for new product entry.
Sales and Marketing:
-Segmentation
-Market Analysis
Insurance
To provide reports containing critical information for the sales staff and management.
Operations:
-Dashboarding
-Automated Reports
- Maintained and automated risk reports
- Responded to Ad-hoc queries requests from hospitals and insurance companies