Our solutions are based not only on finding explanations to understand the past, but also on learning from those experiences to forecast the future, i.e., provide information that can be used dynamically to forecast what will happen every week, every day, every hour, every minute, or even every second. 

Our models do not provide just a one number forecast, but rather a forecast probability function, allowing our customers to quantify risks and adopt sure-fire strategies to reach goals.

 

Some of our solutions have been used to forecast:

 

• Hourly electric consumption, both aggregate and by stations with different time horizons.
• Cash and customers for hundreds of stores and sections of a retail supermarket chain, both daily and weekly.
• Levels of new and churning customers of telephone lines (RTB, ISDN, ADSL, etc.) by central and geographic areas.
• Number of calls to a call center by kind of customer and reasons for calling.
• Price of a square meter by category of the real estate market.
• Level of sales by SKU and optimal discount in the sales period.
• Optimal distribution and allocation of different kinds of periodical publications across thousands of points of sale, on a daily basis.
• The value of clients in the financial sector.