Plan demand in SAP S/4HANA Cloud with AI forecasting and release planning requirements
Demand planning inside SAP S/4HANA Cloud, from sales history to released requirements. AI forecasting runs on a pretrained model, with nothing to train or tune. Every product in scope is forecast every cycle, and new products phase in from a reference item while retiring ones phase out — nothing is missed. Adjustments stay separate from model output, keeping each cycle auditable. Planning discipline for the next level of planned growth.
Overview
Forecast every product without a data science team
Planners produce a statistically sound forecast for the entire product range without hiring or retaining forecasting specialists. A pretrained model does the fitting — nothing to choose, maintain or tune. The forecast is built from live sales history in one place, replacing the monthly scramble to collate spreadsheets, reconcile versions and agree which file is current. Quick to learn, quick to train, and quick to run whenever it is needed.
Rely on a foundation model purpose-built for tabular business data
Large language models learned to read; tabular foundation models learned to predict from rows and columns — the shape business data actually takes. Forecasting runs on TabPFN from Prior Labs, acquired by SAP in 2026, pretrained on millions of synthetic datasets so it brings predictive skill to a series it has never seen. A forecast returns in a single pass, with no per-item fitting, hyperparameters or retraining.
Adjust the AI baseline and release the final forecast
The model produces a baseline; the planner decides the final number. History can be corrected where actuals mislead, and any forecast period can be overridden, with adjustments held separately from model output so the planner's contribution stays visible and measurable. The final forecast is baseline plus adjustment, released as planned independent requirements that drive MRP. It runs as often as the planner needs, in the same few steps.
Details
Solution type
Extensions and Add-ons
Industry
Life Sciences, Consumer Products, Industrial Manufacturing, Retail
Compatibility
Works with
Features
Forecast a defined product scope with a pretrained AI model
Reusable definitions capture what to plan and how to plan it, so a forecast can be repeated without being reconfigured. A pretrained model produces the baseline across the defined scope, including items with no history of their own. The integration is designed for real supply chain planning needs.
Adjust history and forecast without losing the AI baseline
Planners override actuals where history misleads and override any forecast period directly. Model output and manual adjustment are stored as separate figures, so the final forecast is always traceable to both, and the planner's contribution can be measured rather than assumed.
Release the final forecast as planned independent requirements
The final forecast writes directly into SAP S/4HANA® Cloud ERP as planned independent requirements, ready for MRP and supply planning. The release is repeatable, so the plan in the system always reflects the latest agreed view.
