Sramana Mitra: What you’re saying is the backend of the web self-service systems of phone self-service systems have gone from rule-based systems to learning-based systems.
Christopher Connolly: Right. We are at the precipice of that right now. It’s getting to the point of early adoption. It’s a very different way of thinking of how you interact with brands and how brands interact with you.
Sramana Mitra: Do the same thing for me in your employee engagement use case. What is the nugget there?
Christopher Connolly: This is where we really talk about blended AI. Augmented experiences aren’t new but what we’re seeing is the dialog engines that we have are being applied very differently and are being supplemented with machine learning. I’ll give you a concrete example there.
Your best employee who’s been at your organization for 10 years knows a lot of information. They have worked out shortcuts in your systems just by virtue of longevity. By observing what your employees are doing, you are able to learn that information. The machine is able to capture a lot of the essence of what you’re doing there.
The implication is that you have a new employee that starts. In their desktop or phone, they can benefit from the experience of your longest-serving employee. Bots are watching the activity of your new employee and making suggestions in real-time. If you had a flight booking type of situation, people know the tricks. They know where to book, when, and how. People who started yesterday may not know that. That whole employee experience changes a lot.
There are also things like skills assessment of your employees. Being able to feed those results into a machine learning algorithm, you’re allowing the machine to determine who gets what work based on skill set, availability, and business priority. Then supplementing that employee with some cool technology.
Sramana Mitra: Let’s switch to the next line of questioning. What do you see on the horizon? If you were starting a company today, where would you zero in? Where do you see open problems and opportunities for new entrepreneurs?
Christopher Connolly: There’s a problem still around summarization of facts and data that hasn’t been solved well. There’s a lot of textual information in organizations where you have documents, images, PowerPoint presentations that have a lot of information in them that used to be accessed in short form. This is part of the problem of chat bots today. You need that three bullet point summary.
To get to that is something that a human can do very well. That’s still a very hard problem for machines but is something that is extremely necessary and valuable. People have done a decent job at this but democratization of that is an active problem space that needs to be worked on.
Sramana Mitra: Do you have any other ideas you want to talk about?
Christopher Connolly: There’s still a world out there in image processing. There have been many projects out there.
Sramana Mitra: Great. Thank you for your time.