Demand Planning

1. NPI planning – publishing revised forecast for cannibalized items 
   • The application supports publishing the “Revised forecast” of cannibalized items from the NPI plan into demand planning. The revised forecast is populated into a target time series, which is pre-configured in the admin app.
   • Once a revised forecast is published at the aggregate level (i.e., Product-All-All-All), the application automatically disaggregates to the most granular level defined in the target time series configuration. Disaggregation aligns with the target time series allocation strategy configuration (i.e., disaggregate either proportionately, in an equal split, or disaggregate based on alternate time series).
   • The target time series to populate the Revised forecast of cannibalized item is configured in the Admin App as part of a one-time setup for NPI planning.
2. Pause forecast/ Turn off forecast for specific product/customer combinations.
   • The Disallow forecast functionality enables planners to manually control and pause forecast generation for specific planning item combinations within the application. This feature is designed for UI-based control and is independent of source system data. Once enabled, this functionality ensures that forecast generation is blocked for selected planning items.
   • “Data maintenance” mode in the Inspect data screen
     1. Data maintenance” toggle appears in the Inspect Data screen for the specific entity (e.g., product, product-customer, etc.) where the Disallow Forecast functionality has been enabled via the Admin App.
     2. When the toggle is turned on, the application allows the Planners to select items for which the forecast is to be turned off or paused.
     3. By default, the forecast generation will be turned off for the entire future horizon.
     4. Note: The regular data loads will not impact these changes. 
   • “Data maintenance” mode is off – When the data maintenance mode is turned off, the planner can edit other attributes. 
   • The search results on the create forecast screen display an indicator for planning items where forecast generation is disabled.
3. Lock overrides and selective disaggregation logic 
   • A new Lock Disaggregation / Lock Overrides capability is introduced in the planning views. 
     1. Available when the planner right-clicks on a planning time series cell.
     2. The option is shown only if the Lock Disaggregation feature is enabled in the Admin App for that time series 
     3. If the feature is not enabled, the option will not be visible.
     4. The planner can lock the entire horizon or only a few periods.
   • Locking behavior 
     1. Once the planner locks, the application locks editing the selected item, measure, and periods for which the lock is applied.
     2. Business rules do not populate values to locked time series or items.
     3. All related items under the locked item are also locked.
     4. Changes made at higher levels do not cascade to locked items.
   • Unlocking behavior 
     1. Planners can unlock the original item or any one of its relations.
     2. Once unlocked, the planner regains the ability to edit or overwrite values.
   • Partial lock behavior 
     1. The application allows partial lock behavior.
     2. Example: If Product-All-All-All is locked but one of its relations (Product-Customer-Location-All-All-All) is unlocked, the partial lock indicator will be shown at the Product-All-All-All item. The planner can continue editing it. However, the delta changes will reflect only in the Unlock relations. The rest of the locked relations will be unchanged.
   • UI indicators
     1. A lock symbol is shown for locked cells.
     2. A partial lock symbol is shown if some but not all relations are unlocked.
   • Audit trail 
     1. The system captures all lock and unlock events in the audit logs for traceability. 

Replenishment

1. Dependent demand as an independent module 

1. The application now supports the ability to configure and use the Dependent Demand module independently of the Demand Fulfillment module.
2. Implementers can now activate the Dependent Demand along with the Replenishment plan.

2. Running multiple replenishment solvers 

1. The replenishment solver job now includes a performance enhancement that enables it to run efficiently in a multi-tenant environment.
2. The solver can execute replenishment planning across multiple tenants simultaneously without impacting performance.
3. The engine also processes replenishment plans for multiple items in parallel, improving execution speed.

3. Replenishment is in In-memory 

1. Replenishment Planning now runs on in-memory caching instead of relying solely on the SQL database.

1. The engine uses data directly from the in-memory database, resulting in faster performance.
2. Running replenishment plans for multiple items executes more efficiently with reduced processing time.

2. This enhancement also lowers the chance of failures during large-scale planning runs, improving reliability and scalability.

4. Unit of measure in the proposed orders screen 

1. The Proposed Orders screen now populates the Unit of Measure (UoM) based on data provided in the Product Location Master.  If UoM is not available in the Product Location Master, the system automatically uses the value from the Product Master.
2. The usage of the Unit of Measure follows the standard column name configuration, which must be provided in both the Product Location Master and the Product Master. 

Issues resolved

1. NPI – Like item and NPI Item dropdown performances
   1. Issue: In the Create NPI Plan screen, the Like Item Selection drop-down currently loads all items from the Product Master table. In cases where thousands of items exist, the drop-down attempts to load all of them simultaneously. This behavior causes significant performance degradation and often results in the page becoming unresponsive or hanging.
   2. Fix Provided: Improved the dropdown performance. The results will now load quickly without any delay.