Demand Forecast
The Demand Forecast App with Artificial Intelligence is a web application aimed at micro and small companies to help them accurately foreca...
Last verified:
What is Demand Forecast?
AInventory (free) is a browser-based AI demand forecasting app built to generate SKU-level forecasts quickly from uploaded CSV or TXT files. Users provide three core fields — SKU, period, and demand — then choose the data frequency and forecast horizon to run the model. The app compares several forecasting approaches and automatically selects the best one for each SKU based on holdout RMSE.
It is designed to be simple to use, with a clean interface, column mapping from the browser, a downloadable template, and step-by-step guidance inside the app. The workflow includes automatic sorting of series and filling missing dates by SKU, which reduces prep work for planners and analysts. Results are delivered in downloadable CSV files, including forecasts and performance metrics.
The product emphasizes practical forecasting outputs rather than a complex modeling setup. It provides historical charts with forecast overlays for each SKU, plus confidence intervals for time-series methods where applicable. The site presents it as a tool for making inventory and planning decisions with confidence.
It is aimed at planning managers, supply chain managers, demand analysts, operations managers, and SMEs that need reliable forecasting without setting up a heavy technical environment. The website also mentions temporary session processing and offers email support, WhatsApp support, and an option to contact the vendor for isolation or on-premise needs.
Demand Forecast pricing
Pricing model: Free
AInventory is listed as free on the website, with access to the web app included at no cost. The free version is limited to up to 10 SKUs per run and includes limited session usage plus concurrency and stability restrictions. The product includes the CSV template, the in-app step-by-step guide, and unlimited email support. The website says users needing more capacity should contact the vendor, and it also notes that on-premise or isolated installation is available by request.
Demand Forecast pros
- Browser-based, no local installation
- Free version available
- Automatic model selection
- Compares ETS, ARIMA, Prophet, RF, and XGBoost
- Selects best model by SKU
- SKU-level forecasts
- Supports daily, weekly, and monthly data
- 1 to 6-period horizons
- Confidence intervals for time-series models
- CSV and TXT upload support
- Column mapping in the interface
- Automatically fills missing dates by SKU
- Automatically sorts series
- Downloads forecasts.csv
- Downloads metrics.csv
- Historical plus forecast chart by SKU
- Clean interface
- Downloadable template
- Step-by-step in-app guide
- Unlimited email support
Demand Forecast cons
- Free version limited to 10 SKUs per run
- Limited session usage
- Concurrency restrictions
- Stability restrictions
- Only 1 to 6 forecast periods
- Only daily, weekly, or monthly frequencies
- Requires prepared three-column input
- Needs SKU, period, and demand fields
- Temporary session processing only
- Files are not permanently saved
- No explicit enterprise pricing listed
- On-premise or isolation requires direct contact
- Intervals only for applicable time-series models
- Best-model choice depends on holdout RMSE only
- Not positioned as a full planning suite
Frequently asked questions about Demand Forecast
What data does the app require?
The app requires a file with three columns: SKU, PERIOD as a date field, and DEMAND as a numeric field. Users upload the file as CSV or TXT, then map the columns in the interface before running forecasts.
Which forecasting models does it compare?
It compares five approaches: ETS, ARIMA, Prophet, Random Forest, and XGBoost. For each SKU, it evaluates the models and automatically selects the winner based on holdout RMSE.
How long can the forecast horizon be?
The app supports forecast horizons from 1 to 6 periods. It also lets users choose the frequency as daily, weekly, or monthly before running the forecast.
Does it provide confidence intervals?
Yes, the website says it provides confidence intervals at 90%, 95%, and 99% for time-series models. These intervals appear in the forecast output where applicable.
What files can I download after forecasting?
The app provides forecasts.csv and metrics.csv. Forecasts include the forecasted values and intervals where applicable, while metrics includes MAPE, RMSE, MAE, and MASE.
How many SKUs can I process on the free version?
The free version supports up to 10 SKUs per run. The site also notes limited session usage plus concurrency and stability restrictions for the free version.
Does the app handle messy time series data?
Yes, it automatically completes missing dates by SKU and keeps the series sorted. That reduces manual cleanup before modeling.
Who is this tool intended for?
The website targets planning managers, supply chain managers, demand analysts, operations managers, and SMEs. It is positioned for users who want reliable forecasts without setting up a complex technical environment.
Is my file permanently stored?
No, the website says files are processed in a temporary session and are not permanently saved in the database. It also suggests logging out when finished to free resources.
Is there support or an enterprise option?
Yes, the product includes unlimited email support, and the site also lists WhatsApp support. For higher capacity or isolation/on-premise installation, the website says to contact the vendor directly.