Flight Delay Risk Tracker
Completed · Personal project
A personal analytics project focused on turning messy records into a defensible risk classification.
Business problem
Delay risk depends on several interacting factors at once, which makes raw flight records hard to reason about without a structured model.
My role
Sole analyst: data organization, analysis, classification logic, and visualization.
Approach
Explored which factors plausibly influence delays (weather, departure time, distance, and operational conditions) and decided how to represent each.
What I built
Organized the data with pandas, classified delay risk across those factors, and visualized the patterns with Matplotlib. Claude assisted with synthetic data generation and coding guidance.
How I approached it
- pandas for loading, cleaning, and organizing flight records
- Risk-classification logic based on weather, time, and distance factors
- Matplotlib charts comparing patterns across risk categories
- Synthetic data used for learning and testing
Tools used
- Python
- pandas
- Matplotlib
- Google Colab
- Claude
Business value
A personal learning project with no production use. It sharpened my workflow for moving from raw data to a defensible classification and chart.
Skills demonstrated
- Data cleaning
- Exploratory analysis
- Risk classification
- Data visualization
- Analytical problem-solving
What I took from it
Cleaning and modeling assumptions matter as much as the analysis. Documenting those choices made the results easier to trust.