JobDataAI
RBC | Data Science Intern
RBC’s Workforce Management team plans staffing for about 4,000 call centre employees. Those plans run on call-volume forecasts.
What I Built
I moved the team’s call-volume forecasting out of Excel and into Python. My models halved forecast error against the previous method, and I added weekly and daily forecasts alongside the monthly one.
- Heatmap. A weekly executive report of key metrics, generated in Python.
- Course. Designed and taught a Python forecasting course to the team. It set in motion hiring 3+ more data scientists.
Stats
−50%forecast error (RMSE) against the previous Excel method
8 hoursa week saved by the automated heatmap
12colleagues taught
At RBC
Data Science Intern
Campus Ambassador
Data Analyst Intern
The Stack
ModelsProphet, LSTM (PyTorch), SARIMA
ToolsPython, pandas, Git