Analysis Featured

Hire a Robot to Take Notes on Phone Calls

note taking

Business people who spend a lot of time in meetings taking notes may want to meet Clarke, an AI assistant who helps professionals be more productive.

Here is how it works: You invite Clarke to a conferencing call using a phone number and PIN, or merge Clarke with a call you get on your smartphone. Clarke acts as a silent participant and takes notes with a focus on commands like “Clarke, remind me to schedule a followup” or “Clarke, note this down.”

After the meeting, an email summarizing your calls for the day is sent to you and other participants on your list. It can be pushed to collaboration or CRM software and other productivity tools, usually be setting up a custom email address. For the user, summaries and priority lists seamlessly flow once a day via email. In other words, there’s no cloud or local storage necessary.

Clarke’s technology is focused on summarizing calls using machine learning algorithms to add to a list of commands she recognizes and can act upon. As Clarke does the job, vocabulary and commands are added to the database for future use. So, in essence, she learns the communication patterns of her “bosses.”

Where Does Clarke Come From?

Clarke.ai is the brainchild of Christopher Young and Shishir Bankapur. A private company in its early stages, they’ve yet to secure funding. In April 2015, Young, a sales professional, found himself trying to pay attention in meetings while taking notes, and it wasn’t working. Bankapur realized personal calls could also require extensive note taking. The duo realized that many people do not like taking notes or trying to reliably act on what was said in virtual meetings.

“In 20 bullet points, if two are bad, it isn’t a big deal because users are okay with that for now,” Bankapur said. “It allows us to be 95 percent accurate, according to user feedback. Someday, we may offer exhaustive meeting notes, but we’re not a transcription company. It’s a combination of summarized notes and commands you gave to Clarke.”

Clarke triggers approximately 15 bullet points per call. Future developments include email support, centralized billing, and more. Currently in beta, their focus is on getting the service to operate smoothly. She can also create action items when given the proper command, then create a calendar reminder.

“She’ll send a calendar invite,” Bankapur said. “Algorithms teach Clarke to create a map of actions as it learns your preferences.”

Currently, Clarke.ai can interface with U.S. and Canadian conference lines. They are starting to build an in-house system, relying on existing technologies. Competitors are performing keyword frequency tracking instead of using AI technology, which means recorded calls are subject to quality limitations if the keywords aren’t specific enough or thorough. Transcription services simply put voice to text. Clarke.ai seeks to understand the goals of the customer concerning the call with a business-app focus using current AI technology.

The Future of Artificial Intelligence

Clarke.ai are working on an interface with Alexa and Google phone to simplify virtual meeting calls, getting feedback for in-person meetings. With a system like this, Clarke can be dialed with the phone on the table and the caller could say, “This is John, please start taking notes.”

To meet their continuous improvement goals, Clarke.ai needs more users.

“As we add users to the system, we get more data with the larger sample size, Bankapur said. “Technology helps users, users help technology. We also have a little bit of shortage in talent.”

Bankapur and Young see the industry moving toward an AI-as-a-Service ecosystem.

“People are going to have to figure out how to train robots with no data and develop a scalable model,” Bankapur said. “Artificial Intelligence will shake itself out. Everyone is trying to solve speech to text with Siri and Alexa. But that misses the mark on certain things. As AI gets better, there will be more voice-to-text services.”

But with a 20-30 percent error rate, it’s difficult to create workable apps. Clarke.ai is solving that problem in a unique way.

Authors:

Written with Nicki Jacoby.

allen-taylor
Allen Taylor

About the author

Allen Taylor

An award-winning journalist and former newspaper editor, I currently work as a freelance writer/editor through Taylored Content. In addition to editing VisionAR, I edit the daily news digest at Lending-Times. I also serve the FinTech and AR/AI industries with authoritative content in the form of white papers, case studies, blog posts, and other content designed to position innovators as experts in their niches. Also, a published poet and fiction writer.

Add Comment

Click here to post a comment

Your email address will not be published. Required fields are marked *