An Expert Will Be with You Shortly: Giving Chatbots Semantic Smarts

by Daniel Saaristo

“No operators are currently available. Please hold the line. Your call will be answered in the order in which it was received.”

Chances are you’ve heard this before when calling customer service. And chances are it drove you mad, if all you wanted to know was what time a store would open or close.

Estonia-based startup is working to make customer service experiences better and faster by taking customers off hold and giving them chatbots with the semantic smarts to answer questions accurately.

Thank You for Your Patience – It’s About to Pay Off

Chatbots have been around since the 1990s, with generally limited sophistication and success. To make these virtual assistants as human-like in their communication as possible, is using NVIDIA GPU-accelerated natural language processing, machine learning and rich semantics.

Getting it right is crucial because many industries are anxious about putting customers into interactions with artificial intelligence, and suffering a poor customer service experience. A key deficit of many current chatbots is a lack of knowledge of semantic relationships.

Typically, a computer processes free text as a sequence of symbols with no apparent relationships apart from the order in which they appear in a sentence. A human, however, understands the semantics. For example, a person knows that in the sentence “my older brother rides the bike,” the brother is a human being, the bike is an inanimate object, the bike cannot ride the brother and so forth. feeds its virtual assistant what the company calls “semantic enrichment inputs” so that the chatbot can fully understand the messages it receives. Powered by deep learning, the virtual assistant will continue to learn as it receives more messages and applies the semantic knowledge it has been provided with.

“Using NVIDIA’s powerful GPUs, we were able to custom-build a powerful machine learning architecture capable of semantic tagging of both known and unknown tokens based on context,” explains Indrek Vainu, co-founder of AlphaBlues.

My Name is Uku, How Can I Help? is initially focused on bringing its solution into the financial-technology sector. They’ve already seen one of their virtual assistants, “Uku,” employed by LHV, one of the biggest banks in Estonia.

Thanks to the combination of AI and semantic knowledge, Uku can handle simple questions that customers have, like “Where is my nearest branch?” and “How can I transfer money from my account?”

This has huge benefits for customers, who can get answers to their questions much faster. It’s also a boon to bank staff, who can focus on tackling larger, more complex issues for customers.

Inception Startups at GTC Europe is just one of around 2,000 AI startups worldwide that we’re supporting through our Inception program. Through this program, we help accelerate startups by providing them with access to technology, expertise and marketing support.

Many of our Inception startups will be showcased at GTC Europe in October. Get your chance to see how they’re disrupting industries by registering to attend.

Editor’s note, January 2022: This post has been updated to use the new name of the company, formerly called AlphaBlues.