StockSense — Demand Forecasting & Inventory Optimization

Overview

Every retailer and operations team fights the same two-sided problem: order too little and you stock out and lose sales; order too much and cash sits dead on the shelf. StockSense turns your own sales history into forecasts that get the balance right. Upload a simple CSV of date, product, and quantity, and it fits a trend-aware Holt-Winters model per product to project demand for the next several periods, then computes the numbers that actually drive purchasing: average demand, safety stock (from demand variability and your target service level), the reorder point, and a recommended order quantity that covers the forecast horizon. Unlike the AI tools built on large language models, StockSense uses classical time-series statistics that run on your own server — which means there's no API key to manage, no per-forecast cost, and your sales data never leaves your infrastructure.

Key features

• Per-product demand forecasting with trend-aware Holt-Winters smoothing
• Safety stock, reorder point, and recommended order quantity for each product
• Tune lead time and target service level to your business
• Interactive forecast charts and an inventory recommendations table
• No AI API key — runs fully locally with zero per-forecast cost
• Your sales data never leaves your servers
• Simple CSV in — no integration required

Specifications

Product type

Self-hosted software (source code)

License

Perpetual, modify & redistribute in your products

Backend

Python 3.12 · FastAPI · statsmodels

Frontend

Next.js · React (forecast charts)

Method

Holt-Winters exponential smoothing (time-series)

AI API key

Not required — runs fully locally

Requirements

Docker (no GPU, no API key)

Delivery

License key + private source download link