This project is a high-fidelity Forensic Procurement Audit and Performance Dashboard designed for a technical supply chain environment.
It analyzes 300+ simulated procurement transactions (2024–2025) to evaluate supplier performance across:
- Internal house brands (Dayliff)
- International OEMs (Grundfos, Pedrollo)
- Local Ethiopian vendors
The project transforms raw ERP-style procurement data into strategic, decision-ready insights for supply chain leaders.
Technical procurement in East Africa faces three critical “silent killers” of profitability:
Suppliers gradually increase final invoice prices compared to original Purchase Orders (POs), reducing margin visibility.
Unpredictable international shipping lead times disrupt project timelines and risk critical path delays.
Lack of data-driven comparison makes it difficult to justify the value of internal brands (Dayliff) over external suppliers.
This project implements a three-stage analytical pipeline:
- Detects data anomalies (negative lead times)
- Calculates price variance (%) per transaction
- Flags suspicious supplier behavior
A multi-factor supplier ranking system based on:
- Reliability (50%) - On-time delivery performance
- Cost Stability (30%) - Price variance consistency
- Volume Capacity (20%) - Ability to fulfill large orders
Outputs a Supplier Performance Score for ranking and segmentation.
Built using Streamlit, featuring:
- 4-Quadrant Value vs Reliability Matrix
- Supplier tier classification (Strategic / Risky / Tactical / Low Value)
- Real-time filtering and drill-down analysis
Based on 2024–2025 audit findings:
- Dayliff (Global) maintained 0% price variance
- Recommended as a primary hedge against inflation and volatility
- Some international suppliers showed >1.5% price creep
- Recommend activating Right-to-Audit clauses in future contracts
- Local vendors exhibit high short shipment rates
- Recommend implementing:
- Pre-Delivery Inspection (PDI) protocols
- Supplier quality scorecards
| Category | Tools |
|---|---|
| Language | Python 3.10+ |
| Data Wrangling | Pandas, NumPy |
| Visualization | Plotly Express (Dark Theme) |
| Web Framework | Streamlit (Custom CSS) |
| Environment | Jupyter Notebook, VS Code |