Sales BRAINs cuts the development and deployment time for machine-learning-based “propensity-to-action” prediction models from months to hours for increasing targeted marketing results by up to 50% and streamlining the digital customer experience.

Sales BRAINs enables marketers with no statistical knowhow in B2C enterprises such as banks, telecom operators, retail chains and Internet companies to analyze transactional and digital asset data and automatically develop the most sophisticated machine learning models for predicting the propensity of each customer to carry out different actions such as buy/click/respond, and for predicting the impact of each action on the customer’s value.

Sales BRAINs strength is in its capability to leverage the Spark engine for automatically developing and deploying thousands of complex machine learning models, showing superior performance, something which is not possible by manual modeling using other predictive analytics or machine learning tools in the market.

By integrating Sales BRAINs and Real-Time BRAINs the marketers can run real-time triggers and real-time scoring of machine learning models (built using Sales BRAINs) on the digital profile constructed and updated automatically in real time (by Real Time BRAINs). This helps the enterprises to identify real-time sales opportunities that should be treated immediately on digital channels or CRM, leading to increased revenues.

In addition, this enables enterprises to identify digital customer experience problems in real time and treat them promptly, leading to increased customer satisfaction.

Sales BRAINs easily integrates into the company's Sales and Marketing operations process and IT environment (DWH, Campaign Management, marketing automation), providing clear, measurable ROI in months.

Power to Marketing

Sales BRAINs uses a patent-pending technology called Multi-Segment-Modeling for obtaining better lifts as compared to manually-developed models.

The Multi-Segment-Modeling technology can:

Segment automatically the population of customers doing a specific action, the company wants to model, into 20 to 50 sub-segments ; Develop automated, separate prediction models for the same action for each of the sub-segments selected ; Gather scores from each sub-segment's model into one overall score list and rank the customers in that list into percentiles; Compare the lifts in the top percentiles within the ranked list based on the scores of many sub-segments to the lifts among a ranked list based on one model of the entire population, as is done manually today;
Benefit users as comparisons show that lifts among lists based on scores coming from ~50 models on sub-segments are higher by 10%-70% compared to lifts within lists based on scores coming from one model of the whole of the population or of only a few segments.

Using a modern Drag/Drop interface, the application enables users without any statistical background, R programming or SQL coding capabilities to automatically define, develop and deploy complex machine learning models. The results prioritize the different propositions/offers/communications according to the dynamic campaign, marketing and sales strategies, to forecast what will be the next-best-action for each customer.

Harness the Power of Spark Engine

Sales BRAINs™ operates the Spark engine for automatically running thousands of parallel processes using the Spark cluster, enabling high scalability and superior performance. The automatic data management, modeling, machine learning and batch and real-time scoring processes run using Spark SQL Spark Streaming with embedded R components. In this manner, the Spark engine performs all data management tasks, transformations, calculations, modeling and scoring in-memory instead of performing them in a relational DB.

Core Differentiators


Classic tools provided by predictive analytics vendors



Requires professional services of statisticians and BI experts for data management and modeling

Out-of-the-box full business solution for personalized cross/up-sell recommendations

required to develop and deploy hundreds of NBA/NBO models



of model development, deployment and deployment

Thousands of dollars

Less than ten dollars

Quality of predictions


Up to 50% more accurate as compared to manual modeling

Reference population

Models built on whole population

Models built on customers' segments

Model Update
to changing market and business conditions

requires re-development


GUI level

Requires statistical-analytical know-how

Marketing staff friendly

of complex ETL and statistical processes



With other applications

Standalone / Proprietary applications

One-stop-shop for other customer oriented business analytics models using the same G-STAT BRAINs analytical platform


Complex legacy infrastructure with months long integration process, strong bounds and ties

G-STAT lightweight, immediate results, simple licensing, no strings attached