Enterprise automation
Tunas Restocking Decision Support across 126 Branches
A Python-based decision-support workflow that replaces random vehicle restocking with automated recommendations connected to internal ERP data.
- branches covered
- 126Tunas operations
- workflow turnaround
- 3 days → minmanual Excel process automated
- public status
- Demo onlysee disclaimer below
01Problem
Vehicle restocking decisions were previously made without a consistent automated recommendation layer, requiring manual review and producing less systematic allocations.
02Approach
- Connected the production workflow to internal ERP data on the company network.
- Built Python logic to turn inventory and operational signals into repeatable restocking recommendations.
- Delivered a separate public prototype that accepts spreadsheet input to demonstrate the concept without exposing company data or infrastructure.
03Contribution
- Implemented the Python automation and decision-support logic.
- Adapted the workflow to the operational context of 126 Tunas branches.
- Separated the public demonstration from the confidential ERP-connected production system.
04Outcome
- Replaced a random restocking approach with a consistent automated workflow.
- Reduced a manual three-day Excel workflow to a repeatable process completed in minutes.
- Deployed the production implementation inside the company network.
05Limitations
- ⚠️ This is an example / public demo only, NOT the real implementation at Tunas. The operational version runs on the internal company network with real ERP data and is confidential.
- The public Hugging Face Space is a spreadsheet-based prototype, not the production implementation.
- Company data, ERP integration details, source code, and internal screenshots are confidential.