Even as demand for video conferencing continues to grow, the technology is advancing in productive ways. In particular, the latest AI and machine learning capabilities offer the potential for improved meeting experiences and better capture of information.
Today, AI initiatives are coming from endpoint vendors, meeting software vendors and startups offering unique capabilities on their own or in partnership with meeting platform providers.
In an age where so much is being written and talked about regarding AI, it is difficult for IT leaders to discern what is real and what is hype. A 2019 Nemertes Research Group Inc. study of more than 630 end-user organizations, for example, found that just 5.3% were using AI-powered collaboration tools today, though another 43.3% were either evaluating AI or planning to deploy it in the next two years. Often, however, IT leaders weren't clear on exactly what AI means or what value it provides.
What AI-enabled conferencing can do
In the video conferencing space, AI is bringing several enhancements to meeting room experiences:
- Video improvements include active participant framing, virtual backgrounds and resolution upscaling to improve picture quality.
- Audio improvements include active noise cancellation and the ability to filter out sounds from pets, cars, paper shuffling or even food bags.
- Person counting helps facilities managers to learn trends in room use that will enable optimal meeting space provisioning and that will release rooms that are scheduled for a meeting but are not occupied after meeting start.
- Facial recognition enables participants to see the names of meeting participants and learn more about them during the meeting. Facial recognition also allows meeting systems to recognize users as they enter a room to enable fast meeting start. For example, walking into a room could trigger the video conferencing system to say, "Hello John, I see you are here for your 2 p.m. meeting. Shall I start it?"
- Automated notetaking includes transcribing of conversations and capture of action items for later review.
- Real-time translation allows speakers of different languages to easily engage with one another during a meeting in their native tongue.
- Personal voice assistants enable participants to speak to endpoints and apps to start meetings, capture important items and control the meeting experience.
- Chatbots enable easy interaction with conferencing systems for scheduling or feature control. For example, a person could tell an intelligent chatbot to schedule a meeting, giving it the first names of invitees plus the time and date, using AI to make an intelligent guess as to the full identity of the invitees and to find an open time and conference space that is open on all calendars.
- Proximity detection enables the scheduling of meetings at the available conference room nearest to the person requesting the meeting.
AI-enabled conferencing capabilities are now a primary differentiator among vendors. Companies like BlueJeans, Cisco, Google, GoToMeeting, Microsoft, Pexip and Poly, to name a few, all are adding AI-powered capabilities to their endpoints and software, either natively developed or via acquisition of AI startups. Meanwhile, Zoom recently announced a partnership with Otter.ai to deliver capabilities including transcription and action-item capture.
AI capabilities offer significant potential to improve meeting experiences and post-meeting workflows, both of which should lead to improved value from collaboration services. Nemertes' research shows that just 21% of organizations measure value from their collaboration investments. AI's ability to both provide usage analytics, as well as improve meeting experiences, increases the potential for realized cost savings, revenue enhancements and productivity gains by making meetings more intuitive, optimizing use of meeting space, improving in-meeting experience and ensuring the follow-on action items are accurately captured and distributed to meeting participants.
Here's the potential downside to AI-enabled conferencing: More than 40% of Nemertes' recent study participants expressed strong concerns about AI's capture of information, such as biometrics for facial recognition. Vendors must be able to meet information protection requirements such as GDPR and must also assuage fears of loss of control of facial information should a hacker gain access to phones and other applications.
Evaluating the potential for AI to improve video conferencing experiences starts with talking to vendors to understand both what is available today and what is coming down the line -- in the next several months and in the next several years. Look for specific use cases that can address current pain points, especially around areas like meeting start and scheduling, in-meeting performance and meeting-note capture.
Consider deployment of pilots to test the capabilities both of your own vendors, as well as their partners who bring AI-enabled video experiences to your meeting rooms. Finally, be sure to address any concerns or regulatory requirements governing information protection to ensure compliance with governance rules.