Smart-City Research Analytics

Overview

As Data & BI Lead on a Saint Louis University master’s research project, built Python ETL pipelines, forecasting models, and a responsive BI dashboard with live telemetry, scenario testing, and early-warning signals for cross-functional city stakeholders.

Business problem

Where is traffic and PM2.5 pressure concentrated across monitored junctions, how do conditions change over time, and which early-warning signals should operators review first?

Analysis and approach

• Lead backend data systems and BI reporting for the urban analytics prototype
• Engineer Python ETL pipelines with validation for traffic and air-quality feeds
• Develop heuristic predictive models for forecasting
• Design a responsive dashboard with live telemetry and scenario testing

Skills: ETL pipelines · Forecasting · Dashboard design · Data quality validation

Evidence and visuals

• 

  Evidence placeholder — add research dashboard or notebook exports.

Findings and outcome

• Junction-level monitoring is more actionable when traffic and air-quality feeds share consistent location keys.
• Early-warning signals are useful only when data-quality and missingness are visible beside the alert.
• Scenario testing helps stakeholders explore responses without treating forecasts as certainty.

Outcome: Led data systems and BI reporting for an urban analytics prototype monitoring real-time traffic and PM2.5 across 5+ city junctions.

Recommendations

1. 1

   Publish coverage and data-quality notes beside any hotspot or alert visual.

2. 2

   Review forecast error by junction segment before expanding the prototype.

3. 3

   Link the research archive or cleared screenshots once available.