First prototype from IRTEN-ETORGAI project

RTEN  (Smart After-Sale Service for International Customers) is an ETORGAI international research project that began in April 2014 with the participation of 8 Basque companies (I68, Ibermática, Ibercom, NaturalVox, MondragonLingua, AFM, Gashor and Ibarmia) and 4 R&D&I units from the Basque Science and Technology network (I68 I+D+i, i3B, INVEMA and Vicomtech-IK4).

Its objective is to use technology to address the cost problems of after-sale support services for customers with an international dimension, for the particular case of the Basque industrial fabric made up of SMEs that require a high training level to attend to consultations related to capital goods.

The proposed solution considers automating the 1st support tier, facilitating support tiers 2 and 3, grouping the service and automating its customisation and monitoring. To this end, interaction technology and automatic dialogue management, both spoken and written, are being developed, the knowledge of the sector is being modelled, and automatic reasoning and translation technology are being exploited.

The consortium hopes to have deployed a first solution prototype by the end of  2015.

The prototype will include a first version of the automation of the 1st support tier through telephone channels and the website. The aim will firstly be to identify users that access the platform, automatically categorising the reported incident. If the knowledge base has a solution, this will automatically be communicated to the user through the same input channel. To compensate possible voice recognition errors, the figure of the transparent operator will be used, who will transparently transcribe the reported incident for the user.

To facilitate the tasks of 2nd and 3rd tier operators, the prototype will also have a web portal of tickets where the incidents reported by the users will be automatically stored. To provide service to unsolved 1st tier cases, the group operators may consult a knowledge base that will contain a catalogue of the most frequent incidents and their solutions. The knowledge base inputs will be in the original language in which the incident was managed, but automatic translation into other languages may be requested. The automatic translations of the most frequent incidents will be later edited by professional translators.

As a use-case, the first prototype will contemplate the problem of Gashor – a company engaged in the design, manufacture and marketing of bread and pastry ovens, controlled fermentation chambers, silos and cryogenic cells.

In 2016, the consortium will continue working on the second prototype that will include interconnection functionalities to existing customer services in the customer companies, modules to facilitate the optimisation of the 2nd tier support service and automatic reporting functions to customers. Furthermore, the automatic interaction technology, dialogue management, knowledge modelling and automatic translation will be optimised, extending their application to new use-cases.

The benefits expected as a result of the project include the cost reduction of customer support after-sale services, reducing response times, guaranteed at international level, and empowering Basque industrial SMEs and giving them the opportunity to become global