sgeorgiev1993-gif/kingscross-hospitality-ai
This project explores *demand volatility and operational risk* in high-footfall urban districts (e.g. Kings Cross, London).
What's novel
This project explores *demand volatility and operational risk* in high-footfall urban districts (e.g. Kings Cross, London).
Code Analysis
11 files read · 3 roundsA heuristic urban demand intelligence system for Kings Cross, London that ingests public signals (weather, transport, events, venues), computes a busyness score via weighted formulas, detects anomalies against seasonal baselines using z-scores, and renders a single-page dashboard.
Strengths
The hand-rolled ridge regression with cyclic time features is a genuine ML implementation without heavy dependencies, and the anomaly taxonomy with confidence/persistence tracking shows thoughtful domain modeling for explainable urban analytics.
Weaknesses
The core pipeline is a monolithic 17kb script with no tests, significant code duplication across three directory copies, debug fields left in production data, and the anomaly detection produces highly repetitive low-confidence results that suggest the z-score thresholds are poorly calibrated.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
3133
Contributors
2
Files
291
Active weeks
22
Repository
Language
Python
Stars
1
Forks
0
License
—