Connected planning inputs
Consumption, stock, purchase orders and GRN receipts are brought into a common planning workflow.
Domain: Supply chain / Inventory planning
Demand Planning & Inventory Optimization Portal
From fragmented inventory data to a connected planning environment.
CAPABILITIES · Inventory analytics · Replenishment
LIVE · INTERACTIVE APPLICATION
LIVE DEMO
Use the application. Explore the decisions.
Interactive portfolio demo using fully synthetic data.
Open the application in a new tab and return here to explore the architecture.
A 30-SECOND TOUR
- Explore consumption and stock.
- Review replenishment and inventory health.
- Return to this case study for the architecture and implementation context.
01 / THE CHALLENGE
Planning needs the whole picture.
Replenishment decisions depend on more than stock on hand. Planners need to connect historical consumption, open purchases and receiving activity to understand where inventory is at risk and what to order next.
02 / THE SOLUTION
One connected planning environment.
I built a centralized planning portal that combines consumption, inventory, purchasing and forecast data in one environment. The sanitized demonstration substitutes synthetic records at the extraction boundary, preserving the planning logic and the shape of the incoming data.
03 / ARCHITECTURE
From inputs to decisions.
- 01Operational data
- 02Extraction layer
- 03Python / Pandas
- 04Planning & inventory logic
- 05Streamlit application
- 06Planning decisions
04 / WHAT I BUILT
The engineering behind the workflow.
Replenishment workflows
Pharmaceutical and medical replenishment, open-PO follow-up and pharmacy refill planning expose distinct planning tasks.
Inventory health
Inventory-health analysis and consumption-versus-stock views help surface stock risks and guide review.
Reproducible demo boundary
A deterministic synthetic catalog replaces operational inputs, including the reference lists needed by planning logic.
05 / ANALYTICS & BUSINESS LOGIC
Analysis with operational context.
Prioritize inventory
ABC classification supports a structured view of inventory importance.
Read demand signals
Historical consumption and trend detection inform planning review.
Calculate replenishment
Reorder-quantity calculations and allocation logic translate inputs into planning outputs.
06 / TECHNICAL CHALLENGES
Preserve the logic. Isolate the data.
The key demonstration challenge is preserving analytical behavior while replacing every operational extraction. Synthetic inputs retain identical column shapes, allowing Pandas business logic—including classification, trends, reorder calculations and allocation—to run unchanged.
08 / DEMO & PRIVACY
Clear boundaries. Responsible sharing.
Synthetic-data demonstration
Explore replenishment, inventory health and stock analysis in the live application. This case study stays open in its original tab.
Interactive portfolio demo using fully synthetic data.
This portfolio demonstration uses synthetic data and contains no production credentials, patient information, or confidential organizational datasets.