Latest from Pandera Systems

Leaders in Enterprise Decision Automation®.

Categories

Using R Integration leads to Adaptive Forecasting

How one client uses R as an integrated solution with MicroStrategy to create dynamic dashboards that support state-of-the-art predictive analytics

 

This client is a world class distribution company with over $43 Billion in market capital. Currently, they have around 5,000 employees and support a network of approximately 50,000 contractors, who in turn service a huge client base consisting of both business and residential customers alike.

The Challenge

This business already had a long-standing and mature Business Intelligence environment with a large user base. The BI platform supported dynamic ad-hoc reporting environments along with an array of canned reports and dashboards. However, prior to our solution, the system was purely retrospective in nature only allowing users to look at past performance with no forecasting capacity.

Our Objective

One of the main business needs was a flexible tool for forecasting that would be adaptable for various situations. The business has various different product lines and each product line has a different sales pattern across time. As a result, no single predictive model can be used to effectively forecast across different products.

Pandera is focused on cultivating corporate DNA and engineering decision-making environments that mobilize business intelligence and immerse employees in knowledge.

 

Our Strategy & Solution

We used R (an Open-Source and free programming language for statistical analysis) as an integrated solution with MicroStrategy to create dynamic dashboards that support state-of-the-art predictive algorithms. Using this approach, we were able to leverage MicroStrategy’s query engine and front-end API to manage interactions with the EDW and the user, while still using the power of R in the Background to Calculate the metric with a complex ensemble of predictive models.

Besides just utilizing the power of R, our solution to the forecasting dilemma was an ensemble of models that dynamically selected the optimal method based on patterns observed in the data. This dynamic approach in contrast to hard-coding each specific product lines to appropriate forecasting methods not only simplified the solution considerably but also allows for it to be extremely extensible.

The Outcome

This business has adopted the solution and is actively using it to improve revenue predictability and forecast accuracy, driving an improved understanding of order fulfillment needs and staffing, all of which contribute to substantial improvements to customer satisfaction.

Does this sound similar to problems your organization is facing? Do you have general questions about your Business Intelligence infrastructure that you are looking to get answered? Feel free to fill out the form on the right of this page or contact us as it would be our pleasure to offer guidance through these endeavors.

The following two tabs change content below.
Re-engineering decision-making environments to mobilize business intelligence and immerse employees in knowledge, producing peak performance and ongoing success.

Comments