Artificial intelligence (AI) can play an important role in helping organizations improve their unified communications (UC) infrastructure. However, you have to think carefully about your business goals before you start planning a rollout, according to Kevin Kieller, partner, EnableUC.
“You don’t always need AI,” said Kieller, in a presentation on “The Transformative Impacts of Artificial Intelligence on UC” at Enterprise Connect 2019 in Orlando in March. “Anything that delivers better business outcomes has value,” he added, noting that UPS saved 10 million gallons of fuel a year by merely redesigning its delivery routes with the fewest possible few left turns.
However, AI does have the potential to improve decision making, deliver better analytics and reduce costs in workplace collaboration and contact center operations. “Being able to clearly document decisions and action items is one way AI can help the meeting process,” Kieller said. “In the contact center, AI can help agents capture accurate call notes, strengthen the role of virtual agents and provide deeper insights.”
Kieller defined AI as “the capability of a machine to imitate intelligent human behavior,” and discussed four aspects of this rapidly evolving technology:
- Machine learning, which can generate predictive models of behaviors
- Deep learning, in which models can actually train themselves
- Natural language processing (NLP) for understanding speech in the contact center
- Conversational AI, the ability to simulate conversations, such as chatbots
Organizations can tap the storage and processing power of the cloud to harness the power of AI, which can aggregate data from many sources and allows users to test patterns and develop models. “If you capture and store your data, make sure you are gathering sufficient context, such as geography or currency so that you can normalize that data in the future,” Kieller said.
While AI is still in its infancy, Kieller said some of the near-term collaboration applications include speech-to-text making conference calls more searchable, video-based facial recognition, and participant counts for a large meeting room. Other possibilities include speech- or gesture-based controls, like a verbal command to start a meeting and real-time translation for multi-language sessions.
In the contact center, AI can handle routine matters and help live agents make more productive use of their time. For instance, one hotel company was able to reduce call volume to live agents by 40 percent and reduce those agent-handled calls by an average of 60 seconds, Kieller said.
In the longer term, AI’s potential uses include:
- • Human-like avatars that can replace you in a video/audio call, allowing you to listen in while engaged in other tasks
- Engagement, emotion and lie detection via video
- Proactive meeting set-ups by identifying and matching the right attendees
- AI meeting facilitators prompting attendees to stay on time and gather facts in real time
- AI wingmen who support your positions or ask questions for you on a conference call
- In-depth background information for agents with prompts on sentiment analysis or closing a sale
- Meeting efficiency analytics to help managers determine if meetings are effective
As for lessons for today’s organizations, Kieller said, “You don’t need to start from scratch.” Instead, look for tools and models that can be adopted to your needs and plugged into your communication platforms. Stay focused on measurable business outcomes, talk with your legal and marketing teams, and think AI plus your employees – not instead of humans.”