TopN AnalyticsMulti-Source Fleet Allocation & Utilization Tracking
A large EV fleet operator needing daily operational visibility into vehicle allocation across multiple cities.
Where things stood
Knowing how many of a fleet's vehicles are actively allocated and in use — by city and by day — sounds basic, but no single system captured it cleanly.
Allocation signals existed in fragments across charging transaction records, a conversational chatbot flow, and a separate external partner data feed — none alone gave fleet operations a trustworthy daily count.
What we built
Charge transactions, chatbot flow-tracking logs, and an external partner data feed (PostgreSQL) feed a Python multi-source extraction and normalization pipeline.
Vehicle registration-plate prefixes are used to infer city; stitched data loads into BigQuery (raw and aggregated tables) and mirrors to a live-updating Google Sheet for ops-facing use.
Daily distinct-vehicle counts are aggregated by city and allocation type, with a peak-value rollup and source attribution.
How it works
TopN Analytics built a pipeline that extracts allocation signals from all three sources and normalizes them into one common structure, giving fleet operations a single, daily, automatically refreshed view of vehicle allocation by city delivered directly into tools ops teams already use.
What we delivered and learned
The real technical challenge isn't the aggregation math — it's the upstream integration of three structurally different, imperfect data sources into one trustworthy allocation signal, a common but underappreciated challenge in fleet operations data.
What it's built on
Want the full picture?
We're happy to walk through the details, numbers, and trade-offs directly.