My data science journey began with a crisis.
At 9:01am on Monday morning, my phone buzzed with a PagerDuty alert. Instinctively, I pulled up the company dashboard and saw that loan approval rates had doubled overnight.
My heart started pounding.
I was a risk analyst at the time managing a lending portfolio. Approving loans for a higher-risk segment could have a big impact on loss rate.
Bleary-eyed, I dove into the data.
I frantically:
- Wrote SQL queries to pull data,
- Analyzed data in Python trying to spot anomalies,
- Looked for patterns that could explain the sudden spike in approval rates.
I discovered that one specific feature in the credit model had drifted.
Turns out it was a simple timezone error.
This incident marked a turning point in my career, and was the beginning of my journey going from data analyst to data scientist.