This past week, I completed my second internship at Procter & Gamble. While a completely different experience from last summer, I have truly enjoyed my time working virtually on the Product Supply Data Science team.

This summer, I set out to create a representative model of the P&G supply chain in Python. At first, this task may sound simple, but upon further review, one will note that there are dozens of processes working together to ensure excellent efficiency and service from production through delivery. As a result of this complexity, the first few weeks of my summer revolved around learning these intricacies from members of the award-winning P&G product supply team. After gaining confidence in my knowledge, I began planning and then coding my model, which once completed, could take a number of variable settings and simulate resulting outcomes given specified changes in the supply chain, such as lowering safety stock or changing forecasting tools.

The project was a success! After completing the model, I set out to do a run on a scenario the Fabric Care team was interested in testing. I validated the model, ran the different scenarios and was able to deliver sound analysis on the potential effects of the requested scenarios. The team was happy with the results and are now able to consider decisions with more evidence backing them.

A few learnings from this summer:

It is possible to make friends in a completely virtual setting!

• Over the course of the summer, I have been able to meet and grow closer with a number of other interns and new hires. Specifically, I have loved getting to know Michael, Rachel, Lauren, Anna, Alex, Dominic, Ryan, and Brandon in our informal lunches and hangout sessions. We grew through the summer, coping together with uncertainties and building each other up. I look forward meeting all of them in person once it is safe to do so!

There can be incredible value produced by faithfully following the engineering process.

• Last summer, I spent some time on the Data Architecture team. There, I learned the importance of having a defined back-end architecture before setting out on a data project. Before modeling, I defined all of the tables that I would need. Then, I walked through the process and wrote out the definitions for each of the methods that needed to be coded. By thinking everything through beforehand, the actual coding piece of my project was quite straightforward! By working piece by piece, method by method, and knowing exactly how I wanted the end result of the piece to act, I would focus and deliver efficiently and effectively.

Questions are key.

• At the beginning of this internship, the project was a very daunting task. The only way I was able to complete it successfully was by asking questions whenever they came up. The team at P&G was so open and supportive when if came to these questions, so as time went on, any amount of discomfort that might have been present was gone. Clear communication made this summer so much smoother!

Jupyter Lab is really nice.

• When working on projects with multiple scripts, datasets and files to be accessed, the side-by-side directory and code is very helpful. I had only used Jupyter notebook in the past. From now on, its all Jupyter Lab for me!