Loading Forecasts to RI

If you are implementing IPOCS-Demand Forecasting and RI, then you may want to integrate your approved forecast with RI for reporting. The same RI interfaces are used both for IPOCS-Demand Forecasting forecasts and external, non-Oracle forecasts.

If you are pushing the forecast from IPOCS-Demand Forecasting, use a set of POM jobs to copy forecast outputs directly to RI tables. You may also use an ad hoc process to move the forecast data on-demand during the implementation. The process names for the two interfaces are LOAD_PLANFC1_DATA_ADHOC and LOAD_PLANFC2_DATA_ADHOC. These processes should be scheduled to run automatically outside of batch, as this is the preferred way to integrate forecast data (because the volume can be large and you do not want it to impact the nightly batch cycle times).

You must also map the desired forecast run type to the RI application code from the AI Foundation user interface. Once this is done, the integration jobs can detect data to extract. Only one run type can be mapped to RI at this time; use a SKU/store/week run type, unless you have reconfigured the RI interfaces for some other intersection.

Table 3-11 Forecast Extracts from AIF and IPOCS-Demand Forecasting

POM Job Name	Usage Details
W_RTL_PLANFC_PROD1_LC1_T1_FS_SDE_JOB	Extracts the baseline forecast from the AI Foundation table RSE_FCST_DEMAND_DTL_CAL_EXP in the RDX schema.
W_RTL_PLANFC_PROD2_LC2_T2_FS_SDE_JOB	Extracts the final, approved forecast from the AI Foundation view RSE_RDF_APPR_FCST_VW, which contains only the approved forecast export from IPOCS-Demand Forecasting. Requires a run type mapped to the RI app code.

In all of the jobs above, the target table name is in the job name (such as W_RTL_PLANFC_PROD1_LC1_T1_FS as the target for the PLANFC1 extract). Once the data is moved to the staging layer in RI, the source-agnostic fact load jobs will import the data (W_RTL_PLANFC_PROD1_LC1_T1_FS is loaded to W_RTL_PLANFC_PROD1_LC1_T1_F, and so on).

The forecast tables in RI have configurable levels based on your implementation. The default levels are aligned to the standard outputs of AI Foundation and IPOCS-Demand Forecasting (item/location/week) if you do not customize or extend them. If you have modified the IPOCS-Demand Forecasting solution to operate at different levels, then you must also reconfigure the RI interfaces to match. These configurations are in C_ODI_PARAM_VW, accessed from the Control & Tactical Center in AI Foundation. For complete details on the forecast configurations in RI, refer to the Retail Insights Implementation Guide.