Home Press releases Aramix launches a ne ...

Aramix launches a new project for Samsung Italia: news monitoring and sentiment analysis based on AI scoring


Samsung Italia, the national division of the leading international tech brand, relies on Aramix – the business unit of the Datrix Group developing Artificial Intelligence applications for industrial and management processes – for an ambitious AI-based project for assessing the risk of the distribution counterparty.

In this increasingly competitive market scenario, the reputation of companies is a crucial strategic asset; not only determined by those direct activities that every company carries out itself but also by the risks associated with business partners, distributors, resellers, agents, etc.

As a matter of fact, the Aramix data scientist team that signed this project for Samsung Italia mainly focused on promptly intercepting the risk elements of distributors, through intelligent monitoring and sentiment analysis.

“Today this type of activity is a really delicate issue to carry out, and it must be managed with the utmost respect for user privacy and data, matching the technological progress with the applicable legislation – comments Andrea Ronchi, Principal of Aramix. Today we are able to provide data governance and project management fully compliant with the most stringent reference law provisions. Moreover, to collect and analyze data online and activate our intelligent monitoring solutions, we have acquired the license pursuant to art. 134 of the Consolidated Text of Italian Public Safety Laws”.

This solution is aimed at assessing the financial risk of distribution counterparties, which mainly applies to 

  • digitization processes of credit risk definition 
  • calculation models of the financial risk of distribution enriched with alternative data 

This tool is dedicated to all companies that distribute their products through third parties and that have multi-site, cross-sector distribution networks, especially in the Consumer Goods sector.

Thanks to its Natural Language Processing framework for the analysis of alternative data based on a proprietary algorithm, it can :

  • process large volumes of information such as online news, social networks and customer reviews relating to the various points of sale and the company names they belong to;
  • identify risk signals on a daily basis;
  • classify the news and calculate a score to indicate how critical they are
  • classify reviews and calculate the risk associated with each store

Through a daily alerting system, a visualization dashboard and a monthly detailed report, we provide our Client with a systematic analysis, for a constant and updated overview.

A very challenging project since it analyzed the natural language of the users: in fact reviews are often written improperly, with grammatical and semantic errors, as well as mismatching between ratings and textual comments. The algorithm has been trained to also understand colloquial expressions and interpret keywords.

“Thanks to continuous research related to Risk Management, Samsung Electronics Italia internally built a data lake, to organize in a structured way all the information useful for evaluating the quality and consistency of our customers. We worked with the Datrix Group to search the web for weak and strong signs of risk, training the algorithm to carry out a screening and data organization activity as a Risk Manager would do. We believe this is the beginning of an essential journey for the evolution of all insolvency risk analysis tools and more”, concludes Paolo Fontana, Credit Risk Director of Samsung Italia.