Point-of-Sale with Forecasting Analytics for Chelledenn’s Printing Enterprises
A multi-branch point-of-sale, inventory, and online ordering platform that uses time-series forecasting to predict sales and raw-material demand for a printing business in Laguna.
- Role
- Full-Stack Developer (sole developer)
- Client
- Chelledenn’s Printing Enterprises
- Engagement
- Freelance project
- Timeline
- March – December 2025
- Scope
- 3 branches, admin + customer portals
Chelledenn’s Printing Enterprises is a printing and customization business in Pakil and Paete, Laguna, offering tarpaulin printing, sticker decals, laser cutting, T-shirt printing, trophies and plaques, and wood products across three branches. Before this system, sales were recorded by hand and stock was checked manually, which made it difficult to know which branch needed which materials, or how much to order from suppliers.
I designed and built a web platform that centralizes point-of-sale transactions, inventory, purchasing, and branch replenishment, and adds an online storefront where customers can request customized products and track their orders. On top of the transactional data, I implemented forecasting analytics: a SARIMA time-series model predicts each branch’s monthly gross sales, and a weekly demand model recommends how much of each raw material to order.
This was a freelance engagement with the business owner. I handled the entire project as the sole developer, from requirements gathering and database design to the user interface, the forecasting models, and production deployment.
Backend & data
- Designed a normalized MySQL schema covering services, customizable products and their bills of materials, raw materials, suppliers, purchase orders, branch orders, returns, refunds, waste, expenses, and order transactions.
- Built the Laravel backend with role-based access for administrators, branch managers, and staff, two-factor authentication, and an audit trail of every stock transaction.
- Implemented automatic stock deduction from bills of materials at checkout, restock points per branch, and low-stock and expiry notifications.
Forecasting analytics
- Built a Python forecasting pipeline (pandas and statsmodels) that fits a seasonal SARIMA model to each branch’s sales history and predicts the next months of gross sales, shown alongside actual figures on the dashboard.
- Built weekly raw-material demand forecasts that pre-fill purchase orders and branch stock requests, so managers order based on predicted consumption instead of guesswork.
- Ran forecasts as queued background jobs triggered by new transactions, keeping the dashboards current without slowing down the point of sale.
Frontend
- Built the admin, branch, and customer interfaces with Vue 3, Inertia.js, TypeScript, PrimeVue, and Tailwind CSS, including the POS checkout, sales and expense dashboards, branch comparison reports, and Chart.js visualizations.
- Added a customer storefront with product customization requests, cart and online ordering, real-time messaging over Pusher, service ratings, and courier tracking.
- Added Excel and PDF exports for reports, and thermal receipt printing for the POS.
DevOps
- Set up continuous deployment with GitHub Actions: building frontend assets, installing production dependencies, and deploying to Hostinger over SSH on every release.
