The challenge
Operations and logistics teams needed a near-real-time view of the shipments linked to a specific customer account. The data lived in a tracking platform and had to be extracted by hand, then cleaned before it could be used.
Raw exports contained duplicates, numeric status codes nobody could read at a glance, and dates in inconsistent formats.
The solution
A Python application that runs the whole flow and serves the result as an interactive dashboard:
RetrieveDaily API calls across several date types, rotating keys, with automatic retries
CleanDeduplication by tracking number and date normalisation
TranslateNumeric status codes mapped to readable descriptions
DeliverStreamlit dashboard plus filtered Excel export
What the solution includes
Data tableFiltered shipments for the selected account
Pivot tableShipments by ship date and status
Bar chartStatus distribution over time
Pie chartStatus split for any selected day
Excel exportArchive and audit copy of the filtered data
Resilient callsRetry-aware HTTP requests and error handling
Results
- Daily manual extraction replaced by automated retrieval
- One clean record per shipment, with no duplicates
- Statuses readable by anyone, not just by those who know the codes
- A reusable script with Excel output for archive and audit
Stack
Python
Streamlit
Pandas
REST API
Matplotlib
Excel
Client names and confidential figures have been omitted. The images on this page contain no client data.