The COVID-19 pandemic has highlighted the importance of the customer experience for many organizations and illustrated how AI and automation could augment customer service agents.
Since the advent of the coronavirus pandemic, "it's obvious that now our customer experience has gotten more important than ever," Best Buy Health president Deborah DiSanzo said during a Feb. 9 webinar on CX and AI sponsored by speech recognition and AI vendor Nuance.
The online event came as Nuance -- which is heavily invested in the health IT sector with technologies for healthcare providers -- extended its reach into healthcare with the acquisition of startup Saykara, vendor of an automated transcription service for physicians.
Revealed on Feb. 8, the acquisition builds on Nuance's own Dragon Ambient eXperience (DAX) product, which uses ambient sensing technology to listen in to physician-patient conversations.
The system then automatically converts the conversation into structured notes. The notes go through a human review process to ensure accuracy and HIPAA compliance before being sent back to the physician.
Saykara's platform, Kara, appears similar. It also listens in real time to conversations between physicians and patients and transforms the conversation into notes. The process similarly contains human review of the notes. Both Kara and DAX can feed the notes into electronic health records. According to business information aggregator Crunchbase, Saykara had about $7.5 million in funding.
AI and the customer experience
Meanwhile, the pandemic has forced businesses to make their customer service staff work from home, even as demand for many products and services spiked.
"When you look at the variety of channels that any company has to support, it's not affordable any longer to put a human being on every one of those channels," Nuance CTO Joe Petro said.
Automation and AI technologies can augment humans in those channels to do more work or take over some channels, he continued.
Businesses have had to adopt AI technologies quickly to help their customer experience employees keep up with demand. That marked a shift for many enterprises, DiSanzo said.
Joe PetroCTO, Nuance
Then and now
Before the pandemic, many organizations experimented with AI and automation and looked to find "cool" uses for the new technologies, DiSanzo said. Now, those organizations are applying AI to specific business problems.
Now that they have defined the business problems they wish to address, organizations have had an opportunity to adopt AI more quickly and more fully than ever before.
Best Buy Health is a department of retailer Best Buy that sells wearable and in-home health monitoring services and technology that use AI and automation to improve customer experience.
Best Buy Health sells products that include medical wearable devices and phones that include machine learning fall detection models in medical wearables and phones to detect if a customer has fallen.
The models can also use these inputs to determine a person's gait. They can detect subtle changes in a person's gait to predict if they will have a medical problem that could cause them to fall, DiSanzo said.
Some of the healthcare products Best Buy Health sells use AI technology from Nuance, a longtime vendor of voice analytics and natural language processing technology.
Nuance, for example, has products that can identify callers by their voice. The software tools use neural networks to identify a caller's speed of speech and phraseology and the way they hold the phone. This data is combined to form a sort of thumbprint for the caller, enabling Nuance to automatically identify the person the next time they call in.
Nuance enterprise customers can inject that capability into all their channels, enabling conversational phone agents to quickly identify customers and predict their needs after hearing only a few seconds of their voice.
Enterprises need AI help
But even as organizations rush to adopt AI and automation to improve customer experience, they need to remember that it's not always quick or easy to set up models for the first time.
Organizations should try to find a reliable technology partner -- such as Best Buy or Nuance -- that can provide the data, expertise and technology to help create and run AI models, DiSanzo said.
For many organizations without a specialized partner, tasks such as cleaning and curating data can take "quite a long time," she said.
Organizations should also have a clear business problem to solve in mind and focus their efforts on solving that business problem.
According to Petro, organizations often get distracted by the future and predict what they may need years from now. Organizations may then try to create more complex AI models than they need.
This can be costly and can also draw attention away from more immediate needs, he cautioned. He noted a two- or three-step conversational agent, done well, is more valuable than a 27-step model done poorly.
The webinar was moderated by Dan Miller, founder of software advisory and analysis firm Opus Research.