Loading...
office@audascience.com +972-77-8011500

AuDaScience Profile Builder BRAINs for Marketing Machine Learning Modeling

Collect batch and streaming data from DWH, digital sources and legacy systems and AUTOMATICALLY construct and update, in real time, an aggregated Customer Data Platform for a real 360° customer view.

Every data scientist knows that ~80% of data science work required for developing and implementing machine learning (ML) models for marketing (such as cross-sell prediction, churn prediction, lifetime value prediction), is data preparation work.

The algorithms used in the machine learning model are important but the data that goes into the modeling process is key to really improving the prediction level of the model.
This is the reason why data scientists invest the majority of time spent on the development of ML models in creating aggregated, Customer-level attributes and building calculated attributes – such as trend over time and ratios attributes. All for purposes of adding attributes that might potentially have a significant positive statistical impact on the model's prediction.

Unfortunately, data preparation for modeling takes time and requires massive amounts of manual work, even when using ETL tools. Data preparation for updating aggregated data in real time, for the purpose of real time scoring, is even more complex and requires vast experience in RT programming on big data infrastructure.
AuDaScience Profile Builder BRAINs enables the automatic development and update of an aggregated Customer Profile table using engines:

  1. That runs on Spark for creating and updating an aggregated Customer Profile table from raw historic data, on a periodic basis (i.e. daily, monthly, ...).
  2. That runs on Cassandra for creating and updating an aggregated Customer Profile table from streaming data, in real time.

Profile Builder BRAINs enables an analyst to :

  1. Collect customers' raw data (batch or streaming data from digital assets, such as website or mobile app, or transactional systems)
  2. Define the set of aggregations required to run on each raw data (i.e. the maximum amount that a user has entered in the loan amount box on the loans page on the website during the past week).
  3. Using Profile Builder BRAINs engines transforms this logic into a deployment process which runs on batch or RT on the organization's big data & open source environment.

The Customer Profile is updated in real time and enables real time triggering and scoring of machine learning models developed automatically using Sales BRAINs.
In addition, the RT-updated Customer Profile can be exported on a periodic basis (every hour, day, ...) into the data lake or DWH environment for the ongoing use of the organization's analysts.

The main advantages of Profile Builder BRAINs are:

  • Saves time on data preparation work that could take many man-months using coding or ETL and shorten time-to-results from months to hours.
  • Integrated with AuDaScience BRAINs applications for real time scoring of marketing machine learning models, enabling fully automated data science and ML processing from data preparation, through ML modeling and real time scoring, without integration work.
  • Running on the organization's big data and open source environment, assisting in monetizing investment already effected on the establishment of big data environment.

AuDaScience Sales BRAINs™

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

Subject

Classic tools provided by predictive analytics vendors

AuDaScience Sales BRAINs

Configuration

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

Time
required to develop and deploy hundreds of NBA/NBO models

Weeks-Months

Hours

Cost
of model development and deployment

Thousands of dollars

Less than ten dollars

Quality of predictions

Accurate

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

Automatic

GUI level

Requires statistical-analytical know-how

Marketing staff friendly

Operation
of complex ETL and statistical processes

Manual

Automatic

Integration
With other applications

Standalone / Proprietary applications

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

Platform

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

AuDaScience lightweight, immediate results, simple licensing, no strings attached

AuDaScience BRAINs™ turns your big data into a fortune of increased revenues and customer loyalty

AuDaScience presents the most powerful automatic marketing machine learning platform for highly competitive consumer markets

In consumer-services oriented enterprises such as banks, telecom, retail chains, healthcare, insurance and others, the huge database of structured and unstructured customer records and events holds the untapped, precious knowledge regarding customers' future actions. Traditionally, extracting this knowledge and turning it into business tactics required months of effort, modeling and profiling customer segments and product matrices by data scientists and analytics professionals. AuDaScience BRAINs [Big-data Recommendations for Actionable Insights] replaces this with a friendly, automatic and powerful platform that renders more accurate results faster, while requiring dramatically fewer resources.

AuDaScience BRAINs, which runs as an "automatic data scientist", includes hundreds of data management, statistical and machine learning processes. It is based on the R and packaged as sets of automatic applications for answering specific business challenges such as cross-sell, Next Best Action recommendations, churn prediction and Lifetime Value simulation.

These predictive customer-centric marketing models were traditionally developed manually by data scientists. Now AuDaScience BRAINs uses your company's structured and unstructured data to automatically run all data management and statistical processes:

  • Real-time digital asset data extraction using open source trackers and collectors;
  • Automatic real-time construction and updating of a Digital Profile view including hundreds of smart aggregated parameters on customers’ digital experience and activities;
  • Sampling of customers
  • Variables selection
  • Multi-Segment-Modeling segmentation for improved targeting;
  • Machine learning modeling, per each event (product / action / churn event / …) - by different segments;
  • Aggregated segments' scoring calculation;
  • In-sample, out-of-sample and out-of-time validation for each of the models;
  • Deployment of the scoring process on a periodic basis or in real time to real-time-based digital data;
  • Management of all models for quick and easy maintenance.

AuDaScience BRAINs platform includes a set of sales, marketing and customer service optimization applications that provide professional managers with a friendly tool for improving performance of campaigns, retentions and overall customer lifetime wallet share.

AuDaScience Sales BRAINs
AuDaScience Retention BRAINs
AuDaScience Value BRAINs

AuDaScience BRAINs patent-pending Multi-Segment-Modeling technology allows designing and deploying propensity models on customer segments rather than on whole populations. Using this technology, AuDaScience customers enjoy model lifts of up to 50% higher in comparison to manually-developed models using any data mining tool. This leads to higher response rates in outbound and inbound campaigns, augmented revenues from targeted campaigns and visible ROI after only one or two campaigns.

AuDaScience Retention BRAINs™

Customer Retention Optimization – big data analytics for churn prediction and reward optimization

Retention BRAINs provides a unique and innovative approach to customer churn prediction modeling.

Some unhappy customers can get very loud. Many others just silently churn. Within customers' big data hides the priceless information about who is more susceptible to soon migrate to your competition. Within only hours [instead of months], Retention BRAINs turns this data into valuable retention recommendations, enabling you to keep paying customers happy and even win back old customers.

Retention BRAINs is based on the BRAINs [Big-data Recommendations for Actionable Insights] analytical platform, designed and built on many years of statistical and data analysis experience and more specifically, churn prediction modeling and retention optimization modeling for large consumer based organizations.

By using powerful automated machine learning, retention managers with no statistical know-how can easily produce lists of high probability customer churn and win-back which justify retention efforts.

Retention BRAINs enables consumer oriented service providers in highly competitive markets such as banks, insurance companies, telecom operators, media channels, utilities, healthcare organizations, etc., to build and deploy a larger number of churn prediction and customer retention models, by different customer segments, business lines and churn definitions (full churn, partial churn, churn by products).

Retention BRAINs requires fewer resources while providing prediction with more than 50% better accuracy than classic manual modeling practices, thereby preventing potential revenue loss of millions of dollars each year.

Retention BRAINs easily integrates into the company's Customer Service platforms, Retention operations and IT environment (DWH, Campaign Management), providing clear, measurable ROI in months.

Act before they churn

Retention BRAINs makes the design of prediction models for customers of various segments a simple operation. The patent-pending AuDaScience Multi Segment Modeling technology generates dozens of machine learning models in hours, a process that normally requires many months of combined analytics and statistics expertise.

Using a modern Drag/Drop interface, the application enables retention managers to automatically define, develop and deploy many sophisticated statistical customer retention models. The results recommend and prioritize different retention strategies according to the organization’s dynamic service, marketing and sales policies.

Core differentiators

Subject

Classic tools provided by predictive analytics vendors

AuDaScience Retention BRAINs

Configuration

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

Out-of-the-box churn prevention and retention recommendations solution

Time
required to develop and deploy cross/up-sell models [ getting the same lifts]

days-weeks

Hours

Cost
of 1 model development and deployment

Thousands of dollars

Less than ten dollars

Quality of predictions

Accurate

Up to 50% more accurate as compared to manual modeling

Reference population

Models built on whole population

Models built on customers' segments and products

Model Update
to changing market and business conditions

requires re-development every few months

Automatic

GUI level

Statistical-analytical know-how

Service and retention staff friendly

Operation

Of complex ETL and statistical processes

Manual

Automatic

Platform

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

AuDaScience BRAINs Lightweight, immediate results, simple licensing, no strings attached

Automatic Real-Time Digital Actions Prediction - Modeling and Deployment

AuDaScience BRAINs™ next generation, Spark-based, big data, automated machine-leaning platform provides a unique and revolutionary approach to the automatic development, deployment and updating of machine learning models for predicting the propensity of each customer to carry out different actions during the customer’s digital journey.

AuDaScience BRAINs enables a company's marketing analysts to increase their digital (and non-digital) campaign responses and streamline the digital customer experience, by using a powerful, automated, big data machine learning application that can assist them:

  1. Collect in real-time all customers' digital experience activities from digital assets and save raw digital data to a big data environment;
  2. Use Profile Builder BRAINs application to construct a real-time "Digital Customer Profile View" from raw digital big data, including several hundred attributes profiling each customer's activity in the digital assets (websites, mobile app). This digital customer profile is constantly updated in real time, enabling the company to define real-time triggers for near real-time multichannel campaigns and to update machine learning models' scoring in real time;
  3. Based on the real-time customer profile table and existing batch customer profile tables, use Sales BRAINs to automatically develop hundreds of machine learning models for :
    • Predicting the propensity of each customer to perform different actions: buy various products, click on or tap different areas in the browser, respond to different pop-up or push notification messages, etc.;
    • Predicting the impact of the customer's performance of each action on the revenue of the company, with respect to each customer;
    • Real-time scoring, based on the most up-to-date information available on each customer.
  4. Identify the next best offers or actions for each customer, out of hundreds of thousands of possible products, actions or communications, which will maximize the company's revenue, based on the propensity of each customer to perform each action, or expected revenue;
  5. Feed recommendations (in batch/real-time) to campaign management and marketing automation systems for ongoing, integrated operation, which maximizes each customer's lifetime value based on digital big data streaming;

Profile Builder BRAINs enables Marketing analysts to define a business logic of transformations they would want to apply to raw digital data for constructing and updating in real time an aggregated digital profile which includes calculated attributes. Profile Builder BRAINs deploys the business logic using Spark streaming for updating the digital profile in real time.

Sales BRAINs automatically performs all the steps that are currently being applied manually by data scientists, who code or use various machine learning tools to model all the defined actions performed by customers in the digital channels. The models’ score can run on batch or in real-time on the digital profile, which updates in real-time, together with the classic customer profile tables uploaded to memory and joined with the digital profile.

feedback