At DWS 2018, AI pioneers shared lessons from their journeys

IPsoft Digital Workforce SummitArtificial Intelligence (AI) technologies have a history that goes back decades. However, only in the last few years have we seen the rise of mature AI solutions like Amelia and 1Desk that have the ability to reinvent businesses from the ground up. Companies that are embarking on their AI journeys can learn from the experiences of innovative business leaders who are already years into their AI implementations. At Digital Workforce Summit 2018, several of these pioneers shared their AI stories.


Credit Suisse

Suisse’s Head of Cognitive and Digital Services

Jennifer Hewit, Head of Cognitive and Digital Services, Credit Suisse

The goal: Automate the Global Service Desk, the first line of support for 76,000 users spanning 40 countries.

The journey: The global financial services firm onboarded Amelia in June of last year, and went live in December. The company will “switch off the phones” to non-revenue generating users across the bank by the end of June, and Amelia will be the first way users access IT support going forward.

Lesson 1: Think outside of the IT box. “I looked at how I could do this and decided to take an omnichannel approach. With this, we introduced chat, and we introduced Amelia,” said Jennifer Hewit, the Head of Cognitive and Digital Services at Credit Suisse. “This was an interesting exercise for us as we went on a journey to try and figure out, ‘How do we bring an artificial intelligence alive?’”

To refine, optimize, and humanize the user experience (UX), i.e. make it more “alive,” Hewit designed a Cognitive Center of Excellence (CCoE), which included a neuroscientist and linguist in addition to IT and cognitive specialists. This well-rounded team provided Credit Suisse with insights to help them build a user-centric AI experience. “While automation was a priority for me, the user experience was first,” Hewit said. “If I can’t hook my users into the experience of using this new channel for support, they were going to walk away.”

Lesson 2: Put Amelia in front of users early. Amelia learns the nuances of conversation by interacting with people. When she was first introduced, Amelia had a very low effective score and was reading the intent of users at only around 23%. According to Hewitt, putting an early version of Amelia in front of users was done purposefully and turned out to be key to later success.

“[I realized] the more I gave Amelia to IT people, the more she spoke like an IT person,” Hewit said. “We were losing that user experience. The only way I was going to be able to gain that humanized aspect was by getting more and more users to use her… I invited the bank to crowd source her knowledge, crowd source her brain, help me build Amelia.” After crowdsourcing the training of Amelia for five months, Credit Suisse increased intent recognition to 87% on the top-10 call drivers, “which has been huge,” Hewit said.

Liberty Mutual Insurance

Brian Piccolo, AVP of Ecommerce, Liberty Mutual Insurance

The goal: Automating the quote-to-cash process for car insurance.

The journey: Liberty Mutual Insurance set out to combine the personalized consultative experience of a human agent with the speed and efficiency of a web or digital application. “Enter Amelia. Amelia can create an agent-like, personalized shopping experience for customers purchasing insurance online,” said Brian Piccolo, AVP of Ecommerce at Liberty Mutual Insurance. The insurer had an ambitious goal to train Amelia on a complicated customer-facing use case that involves leading customers through 25 to 30 questions in as efficient and comfortable way as possible.


BNP Paribas

Philippe Denis, Chief Anticipation Officer, BNP Paribas

The goal: Create a dynamic digital interface that would reduce the number of calls and queries while improving service quality.

The journey: The global banking company explored AI as early as 2011 and was first introduced to Amelia at a Gartner forum in Barcelona. Last year, they set up Amelia for a single use case to provide 24/7 omnichannel support for shareholders across the globe.

Lesson: Define Amelia’s scope at each step: Amelia has the potential to handle many tasks, and most organizations will choose to start with a limited set of functions and then expand later. The company split this process into three steps:

1) The first step for BNP Paribas was determining which features are most in demand. “How did we define the scope of Amelia? We did data collection,” explained Philippe Denis, the Chief Anticipation Officer with BNP Paribas. This process helped them determine the volume and type of requests that should be automated first.

2) Next, the company worked with clients to define the voice of the client and use that to build the UX. “Client experience is, of course, at the center of this new transformation,” Denis told the DWS audience.

2) Finally, BNP Paribas held workshops and demonstrations with clients to help define and validate future processes as they move forward.

When it comes to deploying new or cutting-edge enterprise technology, it’s never easy to go first. It takes determination, a willingness to learn and improve from failure as well as success, and relentless planning and focus. IPsoft’s AI Pioneers each display these characteristics, allowing companies in financial services and other industries to learn from their innovative example.


More DWS Coverage