Think 2018, the first-of-its-kind IBM event to showcase Cloud, Watson, Data & AI and replacing the short-lived World of Watson show, brought together industry luminaries and technology enthusiasts this week in Las Vegas. The event included the release of IBM Watson Studio, IBM App Connect Enterprise and pronouncements of “democratizing AI.” IBM also unveiled “Watson Assistant” described as “a smart enterprise assistant that brings together AI, cloud and IoT to help businesses enhance brand loyalty and transform customer experiences, while keeping the business and customer data private and secure.”

In essence, Watson Assistant combines the Watson Conversation platform (released July 2016) and Watson Virtual Agent (released October 2016). Watson Conversation includes out-of-the-box tools to build intents and entities in order to automate customer interactions with a natural language interface. Watson Virtual Agent is the cognitive, conversational self-service engine that utilizes those intents to provide answers and take actions across communication channels.

Watson Assistant is the complete package of the two and curated for an industry-tailored market. As Donna Romer, VP of Product of the IBM Watson Platform, explained, Watson Assistant is designed to be used by business or marketing professionals, not developers or engineers. Watson Assistant provides out-of-the-box, pre-trained conversations in five domain areas: retail banking, financial services, telecommunications, insurance, and energy & utilities. The aim is to jumpstart enterprises with a full suite of conversations already geared towards their industry.

IBM’s goal is to empower enterprises with tools to decipher intent with analytic tools and semantic analysis including a “Watson Recommends” capability within Assistant. This functionality attempts to steer conversations back on track if the customer digresses from the original intent, says Romer. Additionally, Watson Assistant enterprise customers will have full access to other suites in the Watson Discovery portfolio, including Discovery Service, Natural Language Understanding and Watson Knowledge Studio.

Autodesk, an IBM flagship customer, has found significant success with their Autodesk Virtual Assistance agent or “Ava”. In a presentation at Think 2018, Rachael Rekart, Director of Machine Assistance with Autodesk, said that since Ava’s launch in February 2017, the intelligent assistant has grown from handling 20,000 customer conversations a month to 100,000, with an average problem resolution rate of 5 minutes and customer satisfaction rating of 85%. Though Autodesk has built Ava on the previous generation of Watson Conversation and Virtual Agent, Rekart expects more success in growing her multi-skilled team to 13-people in order to scale the growth.

Watson Assistant will officially launch in the second quarter of this year, according to Romer. Included will be a service that enables integration of capabilities from third parties to be delivered through Watson, much like Alexa Skills. For example, companies could add a Salesforce CRM “skill” to pull in customer data or to automate specific tasks like authenticating users or filling out address information. This opening up or “democratization of AI” mark a path forward for IBM and its enterprise customers through simpler, smarter enterprise intelligent assistant solutions.

Categories: Conversational Intelligence, Intelligent Assistants, Articles