How can NLP help enhance customer experience?

Whilst the general public might not know what natural language processing (NLP) is, it's one of the areas of AI that most impacts in our every day lives. When you talk to your smart phone or another voice assistant, when you engage with chatbots online, in applications like google translate, and in online search, your answers are delivered almost instantly. This is an incredibly effective means of information discovery and communication, but these services are useless unless they understand what you're saying.

The machine needs to be able to analyse, understand and generate a replication of human speech to provide the optimum user experience. At First Utility, the NLP team are working to enhance customer service improving their conversational agent, using cutting-edge NLP.  The company are currently working on text-based AI assistants by developing a conversational agent called Ask First. Currently, one of the main challenges they're facing is how to better handle the query intent classification, and they're working through this by experimenting with different approaches based on deep learning, word embeddings and reinforcement learning.

Kamel Nebhi, Senior NLP Data Scientist at First Utility who will be joining RE•WORK at the AI Assistant Summit in London this March 15 - 16, spoke with us about his background in NLP and the importance of the work he's doing there. When we spoke with Kamel, he told us that he's 'especially looking forward to meeting the other speakers at the summit and, I will be delighted to share my experience in AI and NLP. I will also be very interested to hear about the different AI use cases in other companies.
'
Kamel first began working in NLP in 2008 during his masters degree, and shared with us how it's been really amazing to see 'how NLP has evolved over time and has reached a substantial applicative maturity during this past decade'. The concept of NLP has been around for quite some time, but the democratization of NLP is due to both technological progress but also historical fact.

Whilst NLP has come on leaps and bounds in recent years, there are still many obstacles when trying to improve customer experience through AI assistants. Kamel and his team have developed a topic modeling based sentiment analysis platform to help monitor conversations over different channels such as email or chat. This can be useful in detecting and predicting customer complaints, tracking sentiment and knowing the overall customer satisfaction, and detecting trends and weak spots. First Utility are using AI for a positive impact 'to provide correct answers to customer requests, but also to assist customer agents in their work by providing recommendations based on the AI assistant.'


Outside of NLP, Kamel shared how he's excited to see customer service improved by other aspects of AI such as computer vision, machine learning and sensors. As these technologies grow and develop, the AI community will 'continue to address concerns related to bias, security and privacy, but also the community needs to work with more transparency and explain how data is used.'


Kamel will explain his work in NLP and his thoughts on the progressions of AI in more detail at the AI Assistant Summit in London next month. You can register here to guarantee your place at the event, and hear more about the rise of AI assistants from experts from Google Assistant, Vodafone, Sage, Cera, University of Cambridge and many more.