DATACTIF® is an adaptive, collaborative Intelligent Open Architecture Platform that transforms heterogeneous information into knowledge. In a more descriptive way : transactional data from operating systems, data of unstructured sources such as Facebook, Twitter, blogs, news sites, financial reports, customer call records, and comments residing in customer relationship management (CRM) systems, etc…
DATACTIF® process all the data of a company, from all sources regardless systems (DB2, Oracle, SQL Server, etc..) and presents comprehensive and analytical results in a usable and understandable way.
The advantage of DATACTIF® is the ability to be adapted to different issues and complex business environment due to its conceptual framework and the usage of machine learning methods and algorithms, subfield of artificial intelligence (neural networks, Self Organized Maps, SVM, Fuzzy logic, text and image mining, etc….).
Machine learning involves adaptive mechanisms that enable computers to learn from experience, learn by example and learn by analogy. There are two learning methods and both are used in DATACTIF® :
VISUALIZATION IN HUMAN LEVEL
Knowledge visualization in accordance to human abilities is the most important step in data modeling. Decision makers have to do more than just optimize their daily work, work that is constantly changing due to market dynamic. Being able to recognize first the changes and to react promptly and correctly is most decisive for both themselves and their businesses. DATACTIF allows easy, immediate and substantive assessment of corporate knowledge through visualization offered by and at all levels:
DATACTIF® offers state of the science functions :
Machine Learning automation that performs training of existing algorithms, for every new data set decided by end user. This function creates new entities in DATACTIF’s data warehouse as well as metadata and updates all related applications.
Clustering. Allows to discover group (clusters) of procedures or clients with common characteristics. Method used, the self-organizing map (SOM), trained under unsupervised learning from input data that produce a two-dimensional discretized representation (i.e. 25 neurons, picture bellow).
Clustering History. Customers Segmentation observed through time, offers a macroscopic point of view on customers evolution in a social and economic context, measuring in same time the efficiency of the Enterprise’s strategy. Clustering History performs clusters comparison between two time periods.
Classification and Prediction. Classifies business procedures, clients transactions, etc… as the most optimum for a company and in same time predicts future clients behavior and business scenarios effectiveness. The intelligent systems used are trained with historical data predicting churn, Life Time Cycle and Life Time Value as well as Response to Promotional Activities. Prediction results can be connected with other important economical factors, such as market share, sales, net profit, growth evolution, etc.
In addition, this tool assists the user in decision making by suggesting optimum actions to be taken in difficult or unknown market conditions.
Association Rules. In the context of a Customer Centric knowledge model, association rules allows to relate clusters with any kind of information provided from both internal such as promotional campaigns evaluation, or external data such as qualitative researches and specially data from social media.
Reporting Tool can combine findings from DATACTIF® and data from other systems (ERP, CRM, etc…) and presents statistics in ad hoc forms offering a holistic view of an enterprises’ available information.
Depends on the sector (retail, fiances, industry) there are additional modules, as the Distribution Network Optimizer that is available in the Retail Version of DATACTIF.
Distribution Network Optimizer. Distribution Network Optimizer based on historical data of existing stores (profitability, surface, employees, facilities, etc), area data such as social, demographic, economic, data concerning competition and customers behavior, Distribution Network Evaluator performs:
For new stores : Evaluation of new site location options, proposal for best emplacement and prediction of future profitability for each option.
For existing stores : Profitability’s Prediction for next years measuring effects such appearance of new competitor or area properties changes, etc…