Machine Learning – Predictive Analytics
Client : A Global Life Insurance Company
Business Challenge: Predicting Return to Work (RTW) duration for its customer’s, who have availed the disability Insurance with the client. Following were the Challenges. This would help the client predicting the RTW Duration for the customer along with fraudulent claims.
Solution: Data was brought into Big Data Ecosystem and thereafter advanced analytics was performed using Spark. Model was trained and at last, Random Forest delivered the best accuracy result.
Big Data – Data Lake Implementation
Client : A HR Consulting Company
Business Challenge: Building a Data Lake in Hadoop, in order to avoid the ETL Challenges with multiple stream data that was hampering the customer service and processing of some time critical use cases related to Payroll, leaves etc.
Solution: Ingestion through Batch mode or Near Real time for different source data and streams into the Data Lake. Advanced analytics on top of this data lake and its integration with multiple other client applications and systems.
Deep Learning – Banner detection
Client : An Advertising and Media Management Company
Business Challenge: Analyzing the visibility of banners, bill boards outside the establishments, and items placement in the shelf etc thereby saving manual effort in checking them on daily basis.
Solution: With the usage of Tensorflow, we built a couple of Deep Learning models that were trained on GPU’s and now in Production at client site.