Stories: Meet Bitext, the technology that is analysing the sentiment behind our online text
Natural Language Processing (NLP) has taken great strides recently. And Bitext Technology is making huge advances in this field. This year, the company funded by the European Innovation Council (EIC) have partnered with Amazon Connect to offer three new services. Bitext will implement a full-fledged conversational Interactive Voice Response in the Amazon Connect contact centre.
Bitext developed a deep linguistic analysis engine that is able to analyse, and process vast amounts of text generated by consumers, companies or employees. The technology can determine the sentiment or tone behind those texts, and it is helping brands such as Salesforce, Ernst & Young, Telefónica or Vodafone to understand what others are saying online about them.
“The future relies on machines understanding humans”
The Madrid-based company employs a scalable and data-driven linguistic methodology to provide a measurable improvement to Natural Language Understanding performance and intent detection. Therefore, the technology can increase up to 3x more virtual agents’ accuracy, thus making them more human-like.
“Bots which understand a language can change the way businesses work and serve millions of customers, 24 hours a day, every day, for any product. That's a game-changer,” said the CEO Antonio Valderrábanos of Bitext during a panel discussion on automation in financial services at FIN/SUM 2018, the largest fintech summit in Japan, held in Tokyo. “The future relies on machines understanding humans”, he added.
In 2020 the company will bring the power of the Multilingual Natural Language Understanding (NLU) engines (powered with Synthetic Training Data) to Amazon Connect. As a first step, Bitext will set up two new services - Instant LexBot and LexBot Booster- available on the AWS Marketplace, with the Custom LexBot projects service also available on demand. With these solutions, the Amazon Connect contact centre will be able to increase the deflection of interactions for voice and chat and reduce time to value.
Bitext received a Phase 2 Grant from the European Innovation Council in 2018. The overall goal of the Phase 2 project was to accelerate bot training and improve the human-machine understanding in this new era of people and machines communication. The total Phase 2 grant was 1.156.391M€.