OptiSupply

Supply chain analytics reducing costs, delays, and inventory.

PythonPandasScikit-learnPuLPPlotly

Impact Achieved

−$200K/yr

15%

Shipping Cost Reduction

−8 days avg

20%

Delay Reduction

−$150K carry

10%

Inventory Optimization

Analytics Components

Supplier Risk Scoring

On-time delivery, lead time variability, and quality metrics combined into composite risk scores.

Low Risk0%

Medium Risk0%

High Risk0%

Lane Optimizer

Linear programming (PuLP) for optimal lane selection balancing cost vs transit time.

Cost per unitOptimized

Transit timeConstrained

CapacityBounded

Inventory Model

Safety stock and reorder point optimization based on demand and lead time variability.

SS = Z × √(LT×σ²ᴅ + μ²ᴅ×σ²ᴸᵀ)

Monthly Analytics Cycle

Data Extraction

Pull orders, shipments, inventory from SAP/TMS/WMS

Transformation

Normalize, calculate KPIs, build rolling averages

Analytics & Modeling

Update risk scores, optimize lanes, refresh inventory recs

Reporting & Action

Executive summary, dashboards, monthly review

500K+

Order lines processed monthly

Key Design Decisions

Interpretability > Complexity

Transparent scoring models (weighted averages) over black-box ML. Stakeholders need to understand and trust the risk scores.

Batch + Exception Alerting

Monthly analytics aligned with decision cycles, plus daily alerts for critical delays (> 5 days late).

What-If > Auto-Optimization

Scenario analysis for human-in-the-loop decisions. Supply chain involves non-quantifiable factors like vendor relationships.

SKU-Level Data, Category-Level Analysis

Maintain granular data for deep-dives, but aggregate for clearer insights and faster queries.

50+

Suppliers Tracked

Countries

500K

Order Lines/Month

Analytics Models