IRS-PM-2023-05-21-ReplenishNow

---

SECTION 1 : PROJECT TITLE

ReplenishNow - demand-driven inventory replenishment recommender

---

SECTION 2 : EXECUTIVE SUMMARY

ReplenishNow is a recommendation engine designed for small-to-medium enterprises seeking to improve the accuracy and efficiency of their inventory restocking strategies. The system implements existing industry methods to forecast inbound stock to balance between safeguard against stock-out (no products to fulfill customer orders) whilst minimizing inventory to avoid over stocking (unnecessary overhead and depreciation costs). Under the hood, ReplenishNow uses a combination of machine reasoning and learning to forecast and monitor the supply and demand levels to make daily adjustments to inventory levels in real time.

---

SECTION 3 : CREDITS / PROJECT CONTRIBUTION

Official Full Name	Student ID	Work Items/Roles	Email
Cao Zengyu	A0214887R	System Design/Architect, Algorithm Developement, Backend Development	[email protected]
​Benjamin Lim	A1234567B	Project Management, UI Development	[email protected]
Toh Zhi Yuan	A0269367L	Data Scientist, Development	[email protected]

---

SECTION 4 : VIDEO OF SYSTEM MODELLING & USE CASE DEMO

ReplenishNow

Note: More reference video presentations here

---

SECTION 5 : USER GUIDE

Refer to appendix <Installation & User Guide> in project report at Github Folder: ProjectReport

[ 1 ] Retool Platform

Go to URL using web browser: ReplenishNow-retool

To start your own app:

1. Register account for Retool
2. Check the application components: App, Retool Database, Workflow, etc.
3. Expose the app ready to use

[ 2 ] Forecasting API

use the source code zip under SystemCode folder to set up in Azure Function

library required: !pip install pycaret

Use the API deployed: POST "https://replenishnowforecast.azurewebsites.net/api/ReplenishNowForecast"

[ 3 ] Forecasting

use the optimiser-cloud-function-source.zip file under SystemCode/replenish_optimiser folder to set up in Google Cloud Function

library required: functions-framework==3.* & ortools

Use the API deployed: POST https://function-optimiser-3phvgqtjuq-as.a.run.app

---

SECTION 6 : PROJECT REPORT

Refer to project report at Github Folder: ProjectReport

Recommended to read the chapters in Project Report:

• Executive Summary
• Market Research
• Project Definition
• System Design (Architecture, data flow, modules design, database schema)
• Demand Forecasting (Theory Study and our methodology)
• Replenishment (Theory Study and our methodology)
• System Implementation & testing
• Project Conclusions: Findings & Recommendation

---

SECTION 7 : MISCELLANEOUS

Refer to Github Folder: Miscellaneous

Data source

• /raw_data_DMC: raw data from DMC2022 competition

Processing files

• inventory.csv, master.csv, sales_order.csv: data being mapped into database tables
• top78Items_aggr_weekly.csv: selected a pool of items from dataset download

Result files

• item_20109_moq.csv: result of the naive replenishment strategy applied on item 20109
• replenish_notification.csv: result of automated decision maker exported

System Design

• system_design.drawio: including all the diagrams

more coming...

---

---

This project is the practical module of the Analytics and Intelligent Systems and Graduate Certificate in Intelligent Reasoning Systems (IRS) series offered by NUS-ISS.

Lecturer: GU Zhan (Sam)

[email protected]