Cognitive

Soul Machines gives AI a human face with the help of Watson

Share this post:

Soul Machines Watson AssistantFrom talking watches to AI-powered financial advisors, a world where people routinely speak to machines is shifting from science fiction to everyday reality.

Artificial intelligence (AI) is no longer in the realm of academic research. It is becoming extremely useful for enterprises, especially in business functions such as customer service. Chatbots are the obvious example. By building an AI-powered chatbot that can respond intelligently to most day-to-day customer service requests, organizations can free up human staff to focus on more complex problems.

But that doesn’t mean it always feels natural. While talking to chatbots can be great to get information or make transactions quickly, a text prompt can’t convey all the subtle nuances of a face-to-face interaction.

If a person is talking to you and you smile, they will register this subconsciously, and there’s a good chance that they will smile back without even thinking about it. These subtle reactions are a big part of how we connect with people on an emotional level, which is valuable in building customer relationships. But how can you make a similar emotional connection with a chatbot?

The human in the machine

At Soul Machines, we solve this exact problem by giving AI a human face, literally. We create sophisticated “artificial humans” that use biology-inspired models of the human brain to speak, move and express themselves just like real people.

To take a simple example, when a customer is talking to one of our artificial humans and smiles into their webcam, our solution registers the image of a happy face and generates the virtual equivalent of dopamine and serotonin in the artificial human’s brain. This causes the artificial human to reciprocate the emotion and smile back at the customer. We adapt tone of voice in a similar way, enabling our artificial humans to communicate in a way that harmonizes with the customer’s mood.

We provide the user-friendly interface that helps AI-powered representatives interact in a realistic, emotionally intelligent way, delivering answers in a manner that makes the customer feel comfortable. But that’s only half the story. To make sure answers are correct and helpful, we also need to give our artificial humans the ability to learn from our clients’ domain-specific data.

Watson brings the knowledge

IBM Watson Assistant does just that. The enterprise-level AI assistant can be trained to answer customer queries for any business while continuously learning and improving its performance.

Many of our clients were already working with IBM Watson to handle customer queries through text-based chatbot interfaces. In each case, we were able to integrate our artificial human with the existing IBM solution to create a new user experience layer. For a company that already has Watson Assistant set up, it only takes eight to 12 weeks to add a Soul Machines artificial human to their solution.

The integration between our platform and IBM Watson Assistant is relatively straightforward to set up. The Watson solution runs in the IBM Cloud—so there’s no infrastructure for us to manage—and provides a simple API that we can call from our application. We send audio of the customer’s voice via this API, and Watson converts it into text, then searches a corpus of knowledge for relevant answers to the customer’s question, ranks the results and returns the top-ranked answer to us.

Meanwhile, our platform is analyzing the audiovisual input for emotional cues from the customer’s tone of voice and their facial micro-expressions. It then converts the answer into modulated, emotionally inflected speech for the artificial human to deliver, matched with appropriately generated facial expressions.

All the evidence suggests that customers enjoy interacting with our artificial humans, and one of our banking clients has found that it can already satisfy 40 percent of their customer queries without any kind of human intervention. We expect that number to climb even higher as the solution continues to learn.

Partnering with Watson

Our experience working with IBM has been great and has boosted our reach as a business. The scale of the IBM Global Services team means they have deep relationships with clients around the world that we can easily build upon.

From a technical perspective, I would recommend IBM Watson Assistant to our customers. We have had success working with Watson, and we’ve now introduced it to several of our newer clients.

As digital transformation gets underway across businesses in every industry, getting the most out of AI solutions will become more and more important. Unlike traditional applications, AI-powered solutions continue to learn and improve over time, so the companies that get AI to market first will have a significant advantage. Their competitors will find it difficult to catch up. We’re proud to be working with IBM to help our clients gain that competitive edge.

Read the case study for more details.

More Cognitive stories

Providing around-the-clock call center support using a cognitive virtual assistant

According to Investopedia, financial technology has grown explosively since the mobile internet and smartphone revolution. Originally referring to computer technology applied to the back office of banks or trading firms, “fintech” now describes a variety of financial activities, from depositing a check with a smartphone to managing investments, generally without the assistance of a person. […]

Continue reading

How to remove the top 3 barriers to AI adoption in business automation

Many businesses are looking to automation to increase productivity, save costs and improve customer and employee experiences. For example, imagine a large insurance company that processes millions of claims a year. Around 60 percent of its claims are automatically processed, but it’d like to automate 85 percent of claims and reduce error rates, reduce costs […]

Continue reading

SPF Private Clients uses AI virtual assistant to support first-time home buyers

In 2013, the UK government launched the Help to Buy initiative to support people struggling to own property. It brought in an influx of mortgage applications to brokers such as SPF Private Clients. In response, we at SPF teamed with EscalateAI to create a portal and AI-powered virtual assistant based on IBM Cloud and Watson […]

Continue reading