You are here:  Home  >  Data Education  >  BI / Data Science News, Articles, & Education  >  BI / Data Science Articles  >  Current Article

Artificial Intelligence: Driving Customer Experience and Data Analytics to New Levels

By   /  August 2, 2017  /  No Comments

artificial intelligenceCustomer experience executives in every industry are intrigued by the idea of using chatbots inspired by Artificial Intelligence (AI) to better service their clients, reduce costs, and increase the possibility for better analytics and other data-related activities.

Many vendors are looking to deliver solutions in this space, one of them being Rulai. Recently, it was recognized by Aragon Research as one of the Hot Vendors in AI Chatbots for 2017 in its report highlighting vendors in the arena with interesting, cutting-edge products, services, or technologies. “The current generation of AI Chatbots leverages significant advances in natural language understanding to provide meaningful, personalized responses that reflect the context of the interaction or conversation,” said Adrian Bowles, VP Research and Lead Analyst for AI at Aragon Research, in a press release. The hot vendors in Artificial Intelligence chatbots make it easier for enterprises to provide more engaging systems for their customers and employees, he added.

What Rulai says differentiates its technology from most others in the space is that there’s no coding necessary to build Artificial Intelligence chatbots. Rather, it offers a point-and-click interface so that customer service execs and teams can themselves create multi-round conversational systems.

The year-old startup was co-founded by CTO Yi Zhang, a tenured professor at the University of California at Santa Cruz and a consultant and technical adviser for several large enterprises in personalized search and recommendation, Natural Language Processing, Machine Learning, data mining, and computational economics. Along with Zhang, the company is staffed by other Machine Learning researchers with award-wining experience in natural language understanding, Deep Learning platforms, dialogue, personalized recommendation and multi-modal user interaction systems.

“In the last few years I’ve turned to the problem of the domain of conversation agents, because it’s very interesting and very technically challenging,” Zhang says. There’s big potential for companies that succeed in helping humans and machines seamlessly interact through conversations, though, with the chatbot market estimated to grow from $703 million in 2016 to over $3 billion by 2021.

Rulai deems its technology, the Rulai Virtual Customer Assistant and Rulai Virtual Agent Assistant, for automating customer and employee interactions as providing an intelligent, highly scalable form of engagement that personalizes the user experience, increases sales, and delivers operational efficiencies.

As VP of business development Jim Diaz explains, the no code, business user/domain expert focused approach Rulai is taking to help brands transform the customer experience is distinctly different from the more expensive, developer focused technologies that big brand companies can afford to invest in to create chatbots, which are highly engineering intensive, built on thousands of pages of code. Rulai sees an inherent advantage in empowering the business user, allowing that domain expert the ability to imprint their best practices while building AI Chatbot use cases within the Interaction Design Console, seamlessly leveraging their 1st party data as necessary, then relying upon Rulai’s AI system to understand customer intent and enable multi-round conversations that result in deterministic, defined resolutions or agent escalations.

Rulai’s approach departs from most “legacy” solutions in the technology behind its assistants: MITIS, which stands for Mixed Tasks and Initiatives System, aims to expand the notion of AI-evolved chatbots. Today many chatbots are aimed at solving Tier 1 issues, being able to comprehend and respond to simple requests – lost passwords, for instance, Diaz says. “Those are the basic things you can write a few rules around linear conversation – ask one question, get an answer. It’s like basic FAQs can be applied to these rules-based chatbots,” he says. By creating the capability to scale up use cases to more complex, multi-round issues utilizing natural language in the way a customer would converse with another human being, MITIS’ technology represents a big advance over that.

“What we do is create a way for the business user to come in, without coding or programming, and be able to create the dialogue workflow for multi-round, multi-initiative tasks so that virtual assistants can let someone cancel one reservation and then pivot to book a new trip, for example,” Diaz says.

That adds a lot more value for resolving customer issues than simply letting them reset a password.

Not only that, but the technology can help live agents when conversations do need to be escalated to them, populating key data fields in real-time within the context of the conversation to support ongoing recommendations and create a more efficient process for the company and the customer.

Behind the Customer Engagements

When it comes to its interaction management console for creating, configuring, and modifying its virtual assistants, the transparent self-service process involves a drag-and-drop user interface. Remember, Diaz notes, those working on this are not engineers versed in the space (those resources are rare even in large companies).

“The key use cases for us are to find customer experience leaders that want to take control of this, who don’t want to outsource the transformation of their customer experience to a black box either built by an engineering team or a third-party vendor,” he says.

For conversational agents, especially in the short term, Zhang says, “you need to integrate domain knowledge from managers instead of having software engineers who don’t know the domain build the bot.”

In their target enterprise market, call center metrics are already there to provide unlabeled or labeled customer data (which can be catalogued into its pre-defined entity types or custom-defined entities).

“One of the interesting things about the technology is that we try to make it really easy for enterprises to give us access to data that we can get started with – things like FAQs and knowledge-based training data that we can ingest relatively easily and organize, categorize and present in the console so the business user can play conductor,” Diaz says. “It’s like the instruments have all been built and the musicians are in their seats and the customer experience leader job now is to  conduct the orchestra, writing high-level rules around it.”

The system starts with leveraging its Deep Learning method for general domain models that understand common ways of saying or spelling things, for instance, and then engages with new specific domain knowledge atop that. Adaptive learning continues as the chatbot goes into action; when it launches and talks with humans it constantly updates its knowledge from real human-to-human interaction. It also learns from human-to-machine conversation data, ensuring that its smarts grow.

The intelligent, text-based conversational interface simulates human conversation with the ability to recognize intent, provide answers, complete multiple questions and tasks, and handle interruptions. Chatbots are constantly updating their default confidence levels in understanding and updating, and directing a customer to a human agent when confidence sinks past a pre-set threshold.

The adaptive learning analytics, Zhang says, involves conversational agents themselves estimating their needs to learn further. “With our consoles you can review what the bot thinks it needs to learn and the manager can tell it to get more information to resolve an issue,” she says.

A recent user survey conducted by Rulai shows that by using adaptive learning and the console, customer experience successes doubled and tripled within one day, far faster than it would take if a whole process involving managers and engineers and writing code were required. With adaptive learning, Diaz says, the notion of maintenance and heavy lifting that happens with typical rules-based systems goes away.

Diaz explains that competing on customer experience and costs is just part of a picture whose central focus is really about getting first contact resolution. “How at the drop of a pin do you resolve an issue for someone coming into a live environment who wants to get something done?” he says.
Photo Credit: Elnur/Shutterstock.com

About the author

Jennifer Zaino is a New York-based freelance writer specializing in business and technology journalism. She has been an executive editor at leading technology publications, including InformationWeek, where she spearheaded an award-winning news section, and Network Computing, where she helped develop online content strategies including review exclusives and analyst reports. Her freelance credentials include being a regular contributor of original content to The Semantic Web Blog; acting as a contributing writer to RFID Journal; and serving as executive editor at the Smart Architect Smart Enterprise Exchange group. Her work also has appeared in publications and on web sites including EdTech (K-12 and Higher Ed), Ingram Micro Channel Advisor, The CMO Site, and Federal Computer Week.

You might also like...

Thinking Inside the Box: How to Audit an AI

Read More →