Meeting that challenge has become an emerging and increasingly crowded market called conversational AI. Big Tech companies such as Microsoft, Amazon, Google and Oracle have offerings, as do smaller companies and start-ups, including Kore.ai, Omilia, Rasa, Senseforth.ai, Verint and Yellow.ai.
The suppliers provide software tools that companies then adapt and train on their own data.
This year, the enterprise market for virtual assistants, or chatbots, will grow 15 percent to more than $7 billion, according to a forecast by Gartner. Some of those bots are designed to help employees, but most are for customer service.
No company has taken a more humbling and educational journey into its chatbot technology than IBM. After his Watson supercomputer triumphed over human champions in the TV game show “Jeopardy!” about ten years ago, IBM began to apply Watson’s natural language processing to other areas. An early focus was cancer diagnosis and treatment, and IBM called healthcare its “moonshot.”
In January, after years of struggling, IBM announced the sale of its Watson Health business to a private equity firm. A few days later, Gartner rated IBM’s Watson Assistant a “leader” in conversational AI for businesses. Watson has gone from cancer moonshots to customer service chatbots.
Today, Watson Assistant is a success story for IBM among its remaining AI products, including software for exploring data and automating business tasks. Watson Assistant has evolved over the years and is steadily being refined and improved. IBM quickly discovered that a rigid question-and-answer approach, while ideal for a game show, was too limited and inflexible in customer service environments.
“The real world has opened our eyes,” said Aya Soffer, vice president of AI technologies at IBM Research.
The starting point for improvement, said Dr. Soffer, was a deeper understanding of what goes on in call centers, partnering with other companies to mine and analyze many thousands of phone calls between customers and human agents. In dialogues, for example, she tracked what questions and what follow-ups led to solving a customer’s problem, she said, and what were the telltale signs of “bad conversations.”
