Import Forecasts from AI Foundation

Forecasts can be generated from AI Foundation (AIF) and imported to APCS using RAP integration. AI Foundation can generate different levels of forecasts as needed by different levels of plans. It generates both Pre-Season forecasts (using the Auto-ES Forecast method) and In-Season Forecasts (using the Bayesian Forecast Method). AI Foundation directly gets the actuals through RAP integration. Job flows in AI Foundation need to be scheduled to generate the forecast and import the same to APCS. For more details, see the Oracle Retail Analytics and Planning Integration Implementation Guide.

In order to get forecasts from AI Foundation, during implementation, some initial setups need to be done in the AI Foundation platform. For more details, see the Oracle Retail Analytics and Planning Integration Implementation Guide.

Note:

The AI Foundation default Location Hierarchy does not contain the Planning Channel. Customers need to use the Alternate Hierarchy setup which uses the Planning Channel as an alternate hierarchy and create forecasts at the Planning Channel level as needed by MFP GA. For more details about creating forecasts at the Planning Channel level, see the AI Foundation documentation set.

The following table shows the interface table column details from AI Foundation in RDX used for the interface.

Interface Name: RSE_FCST_DEMAND_EXP

Table Column Data Type Comments

RUN_ID

Number(10)

The export Run ID as obtained from the RAP_INTF_UTIL.

CAL_HIER_LEVEL

Varchar2(30)

The calendar level data is for Fiscal Year, Fiscal Quarter, Fiscal Period, Fiscal Week, and Fiscal Day.

LOC_HIER_LEVEL

Varchar2(30)

The location hierarchy level data is for COMPANY, CHAIN, AREA, REGION, DISTRICT, LOCATION, and CHANNEL.

PROD_HIER_LEVEL

Varchar2(30)

The product hierarchy level the data is for CMP, DIV, GRP, DEPT, CLS, SBC, STYLE, STYLE_COLOR, and ITEM.

FCST_DATE_FROM

Date

The start date which the forecast is for.

LOC_EXT_KEY

Varchar2(80)

The external id of the location. It will use the integration ids as provided to RI (preferably the RMS id, and not an integration id such as AREA~123).

PROD_EXT_KEY

Varchar2(80)

The external id of the product hierarchy. It will use the integration ids as provided to RI (preferably the RMS id, and not an integration id such as CLS~123~456~789).

CUSTSEG_EXT_KEY

Varchar2(80)

The external id of the customer segment. It will be NULL if not applicable.

FCST_TYPE

Varchar2(20)

The type of forecast. PI, NPI (PI=Plan Influenced, PI = Non Plan Influenced)

REG_SLS_QTY

Number(38,20)

Regular Sales Units

REG_SLS_AMT

Number(38,20)

Regular Sales Amount

PR_SLS_QTY

Number(38,20)

Promo Sales Units

PR_SLS_AMT

Number(38,20)

Promo Sales Amount

CLR_SLS_QTY

Number(38,20)

Clearance Sales Units

CLR_SLS_AMT

Number(38,20)

Clearance Sales Amount

REG_PR_SLS_QTY

Number(38,20)

Regular and Promo Sales Units

REG_PR_SLS_AMT

Number(38,20)

Regular and Promo Sales Amount

SLS_QTY

Number(38,20)

Total Sales Units

SLS_AMT

Number(38,20)

Total Sales Amount

RET_QTY

Number(38,20)

Return Units

RET_AMT

Number(38,20)

Return Amount

The same Interface table contains the forecast data for different levels of plans differentiated by _LEVEL columns within the interface. The single interface run pulls data for different levels of forecasts which are pre-configured. Customers using non-template versions, if using different levels of plans, can use the supported levels in AI Foundation to generate forecasts. The following sections provide the default levels of forecasts exported for the APCS template version and their mappings.

Item Level Forecasts Mapping

The following table shows the mapping for pre-season and in-season Item Level Forecasts.

Table Column Mapping for Pre-Season (MPP) Mapping for In-Season (MPI)

CAL_HIER_LEVEL

Fiscal Week

Fiscal Week

LOC_HIER_LEVEL

LOCATION

LOCATION

PROD_HIER_LEVEL

ITEM

ITEM

FCST_DATE_FROM

WEEK

WEEK

LOC_EXT_KEY

STOR

STOR

PROD_EXT_KEY

SKU

SKU

CUSTSEG_EXT_KEY

NULL

NULL

FCST_TYPE

NPI

PI

REG_PR_SLS_QTY

FCTYFCPMU

FCTYFCIMU

REG_PR_SLS_AMT

FCTYFCPMR

FCTYFCIMR

Sub-Class Level Forecasts Mapping

The following table shows the mapping for pre-season Sub-Class Level Forecasts.

Table Column Mapping for Pre-Season (MTP)

CAL_HIER_LEVEL

Fiscal Week

LOC_HIER_LEVEL

LOCATION

PROD_HIER_LEVEL

SBC

FCST_DATE_FROM

WEEK

LOC_EXT_KEY

STOR

PROD_EXT_KEY

SCLS

CUSTSEG_EXT_KEY

NULL

FCST_TYPE

NPI

REG_PR_SLS_QTY

FCDVSlS1U

REG_PR_SLS_AMT

FCDVSLS1R