Credit Project Parallel Processing
Credit Project parallel processing clubbed with merging of ADF/ logic apps and better process to calculate the credits USD provides the scalable solution in which we can analyze the credit project in excess to more then 1M input records and using lesser number of ADF/logic apps which results less maintenance cost and more code re-usability and providing more accurate refund which saves 0.5% for Microsoft from total refund amount.
Business Problem
- Unable to process big credit projects having more then 20K input records.
- Unable to process credit projects spanning more then 3 months data.
- Accuracy of refund is less in case of offer change within month.
- Maintenance cost is high since we are creating separate logic apps for each channel.
- Development cost is high for new channel
Solution
- Split the input file into multiple parts and process the splitted files and then recombine them to produce output sheet.
- Merging multiple ADF/Logic Apps to increase code re-usability and handling uncommon scenarios with the help of wrapper SP’s and switch/If Conditions.
- We have only to develop uncommon part for any new channel
- Taking better approach for credit amount calculation in case of subscription change within same month.
- Better mechanism implemented for Error Handling
Why is the solution unique
With the help of this idea, we can now process large amount of input records over longer time span hence lot of manual efforts to calculate and validate credit projects is saved. Now we are calculating with new mechanism that is saving cost(approx. 0.5) for Microsoft.
- Adding new channel will take less time.
- Maintenance cost is lower to mange less ADF/logic apps.
- Able to process big credit projects and credit project with bigger time span.
- Better mechanism for calculating Credit Amount in case of subscription change within month.
Features/BP diagram of the solution
The primary advantage of the solution is tool is now able to process more then 1m records saving a lot of manual efforts and decreasing development and maintenance cost of system and provide better mechanism for calculating credit amount.
