Gas Notification Pipeline
Developer · 2024–2025
⚡ 92% reduction in manual tracking time
Problem
Every hour, the gas engineering team was opening a SQL IDE, writing queries, running them across all their meters, and reviewing results by hand. That was the process for monitoring 25+ high-volume industrial customers. It was time-consuming, easy to miss things, and any incident only got caught on the next scheduled check.
What I Built
- 01Replaced the manual query cycle entirely with an automated hourly pipeline. Engineers don't open SQL to check meter health anymore — the system checks for them and reports back when something looks wrong.
- 02Detection logic uses SQL CTEs and window functions to compare each reading against a rolling baseline. It flags anything faulty or anomalous across all monitored meters in a single pass.
- 03When something looks off, an automated email goes out immediately. The team gets notified before the issue compounds rather than finding it an hour later on the next manual cycle.
- 04A monitoring dashboard tracks pipeline runs, alert history, and meter health over time so the team can see patterns and verify the system is working as expected.
Hourly meter data ingestion → SQL CTEs + window functions (anomaly detection) → Threshold + pattern-based fault classification → Automated alert dispatch → gas engineering team → Monitoring dashboard (alert history + pipeline health)
Results
92% reduction in manual tracking time, measured against the old process of opening SQL, writing queries, and reviewing results every hour. Detection shifted from reactive to proactive. The team now gets notified within minutes of an anomaly instead of catching it on the next scheduled check.
Stack
SQLCTEsWindow FunctionsPythonETL
Next Project
AlphaPulse→