The Client is a leading Spare Part Manufacturing Company for Commercial Vehicles and Trucks, with a sizeable market share in the country. The Company earns a sizable sum of annual revenue amounting to $2 Billion but loses a fair share of its Revenue to claims made by Fraud claimants. In the past year, it faced losses totaling to $500 Million to pay out such fraudulent claimants.
Which Product comes back with repeated complaints.
On which particular routes vehicles break down most number of times.
We developed a Big Data analytics platform using Golang , Python, and Hadoop to pull data from Terra Data. The Platform can produce Historical Analytics based on Historical Data, and at the same time, it offers Predictive Analytics based on The Hadoop Data storage. The client has a Data volume turnover of 2PB annually.
This Data is in the form of Structured as well as Unstructured Data which is sourced from Sales, Marketing, Purchase, Claims paid/made, Dealers/Customer Information, GPS information and much more.
Deciphering this massive amount of data and previous GPS information the platform can determine:p>
Historical Data Analytics
- Clients who have made false claims
- Reasons for breakdown of vehicles.
- Which product reported most complains
- Which spare part or product will report most number of breakdowns
- Which route will cause most breakdowns
The predictive/ Big Data Analytics platform thus provides the Client with crucial information. This prediction aids them in taking accurate actions to avoid fraudulent claims.
The cluster for creating this platform includes:
- 13 Computers
- 1 is a master node with 32GB RAM, Hard Disk 2GB, Octa-core Processor
- 12 Data Nodes with 12 GB RAM and Quad Core Processors
- 937 Vendors were found to be making False Claims
- Total claims against poor quality products have reduced by 50%, and profit margin has increased by $250 million
- $ 250 Million was saved from being paid to false claimants
- There was a Cost Saving of 50%
The Spare Part Manufacturing Company has benefited immensely from this solution.