Ashish Adhikari

A Python-based ETL architecture. I was the back-end developer tasked with simplifying the ETL process. The original process required creating a new Python script for new extracts or load procedures. Working with other team members, we modified the entire ETL procedure to work on the existing libraries while removing the requirement of having a separate script for each process. The entire ETL process was parameterized such that changing any existing procedure or creating new ones could be done based on parameters. By doing this, not only was the redundancy created by having similar scripts reduced, but it also gave an opportunity for people without any knowledge of Python to create new ETL scripts.

A Bash script along with an Oracle BIP report. Working alongside a business analyst and client representative, my task was to develop a Bash script to automate the inventory replenishment process. This project required us to develop a script based on client input parameters to scan the inventory, recent sales trends, and currently ordered quantity of any given product within the company to recommend whether a new order would need to be created in the next replenishment cycle and if so, how much should be ordered. Similarly, a report was generated to provide clients with information and suggest if any of their products would be a good fit for this replenishment model. We developed a mathematical formula that would help predict how the automatic product replenishment would behave for any particular item. Based on this prediction, we added further modules to recommend to the client if the given item would be a good fit for the new replenishment model or not. All this information would be provided to the client in an Oracle BIP report.