Data Science Intern, Winter 2027

Tidewell reads pressure and flow sensors in municipal water pipes and flags leaks before they reach the street. Our anomaly models catch about 70% of leaks within a day, and we want that higher.

As our winter data science intern, you'll work on one research question: can pressure changes caused by a burst pipe be told apart from those caused by a fire hydrant being opened? You'll start by labelling historical events with our field team, then build and evaluate a classifier against the current rules. Expect Python, pandas, scikit-learn and a Timescale database with five years of readings.

You'll work 35 hours a week with a daily overlap from 15:00 to 17:00 UTC. Your mentor will hold a weekly 1:1, and you'll present your findings to the whole company at the end.