Sensible Machine Learning (ML)

No. Use of Sensible ML does not require in-house data scientists or someone who understands machine learning.

Sensible ML is designed to provide the most sophisticated forecast from the historical time series data supplied
customers. Here are a few examples of how Sensible ML can be applied across various planning processes
and use cases:

• To support strategic planning processes, many organizations create 3 or 5-year (or more) forecasts to provide greater clarity of vision and roadmap to support intent of the company, identify strategic choices, ensure clear direction and drive competitive growth.
• To help model strategic plans, organizations may share 5 years of monthly history or 60 historical data points with Sensible ML.
  • Using these 60 data historical data points, Sensible ML will create a statistical forecast model using a single variable (e.g., univariate model) for deployment across OneStream.
• To support Annual Operating Planning (AOP) or forecasting processes, organizations typically require their lines of business to translate “top-down” financial goals into granular, “bottom-up” monthly plans across product categories, sales channels and customers which can result in hundreds of forecasts.
• To help model granular bottom-up plans for the AOP or forecasting, organizations may share 150-250 historical monthly data points with Sensible ML.
  • Using these 150-250 historical monthly data points, Sensible ML is capable of:
    • Ingest and enrich the source data with external macro-economic data.
    • Automatically generate features and select the most influential variables for modeling.
    • Automatically compare and contrast predictive models such as ARIMA / SARIMA methods in the model arena.
    • Automatically cycle through and surface the highest performing model types across all forecast targets for deployment across OneStream.
• To support daily or weekly Demand Planning and/or Sales & Operations Planning (S&OP), organizations typically require their demand planners, business analysts and/or finance business partners to create granular, product and location level forecasts aimed to guide tactical staffing, procurement, logistics and inventory management decisions.
• To help model daily or weekly demand plans, organizations may share 250+ historical data points with Sensible ML.
• Using 250+ historical data points, Sensible ML will unlock the full potential of advanced ML modeling capabilities including:
  • Ingest and enrich the source data with external macro-economic data.
  • Incorporating unique business intuition related to events, promotions, and holidays that are known to influence forecasts through Sensible ML’s event builder.
  • Automatically generate features and selecting the most influential variables for modeling.
  • Automatically compare and contrasting ML methods in the model arena.
  • Automatically cycle through and surface the highest performing model types across all forecast targets for deployment across OneStream.

It could be if you are a OneStream SaaS customer in retail or CPG manufacturing with a use case that matches the current capabilities. Sensible ML provides time series forecasting capabilities with a strong focus on daily, weekly and monthly demand planning and other use cases for FP&A and operational teams across manufacturing, retail, consumer product goods (CPG) and other industries. Based on the amount of historical data provided, Sensible ML can provide a range of predictive or ML models that are ready to be leveraged across your organization and in virtually every industry.