David Shim is the Co-Founder and CEO of Read AI, which combines AI and analytics to make hybrid meetings more effective. On average Read identifies 23 meeting hours a month that could be eliminated per employee. Read’s mission is to make every human interaction meaningfully better, smarter, and happier starting with the more than 500 million people that video conference daily.
Prior to Read, David was the CEO of Foursquare, the location layer of the internet. While at Foursquare, the company exceeded $150MM in annual revenue, achieved profitability, and acquired its two largest competitors, Placed and Factual, to create the de facto leader in location. David joined Foursquare through Snapchat, which had acquired his first startup, Placed for more than $175MM in 2019.
In this exclusive interview, Co-Founder and CEO of Read.ai, David Shim, discusses how his company uses AI in transformative ways, how you can revolutionize your meetings, and the implications of ChatGPT. Don’t miss this!
What are Read’s products exactly and how will they allow organizations to gain more value from meetings?
Here are a couple of real-life examples from organizations using Read:
1. Applying Read Meeting Summary to all meetings. A technology startup has included Read on every meeting with the idea that if the meeting is important enough to have, it should have meeting notes. This has created a library of meeting summaries, topics, action items, and key questions that act as an agenda for all recurring meetings, making meetings more efficient and engaging (as people don’t have to worry about note taking).
2. A 400 person technology company used Read Recommendations to eliminate 8 hours of meetings a month per employee, giving back over 3,000 hours a month to their team. On an annualized basis, Read is set to return $2.2MM in time back to the company and its employees.
3. One of the strongest pilot clients of Read Meeting Summaries are venture capitalists. With back-to-back pitches, updates, board meetings, the ability to have a virtual assistant generate a summary of all their meetings along with follow-up items allows VCs to effectively scale. We’re also seeing something similar with executive assistants, where they are able to use Read’s AI to enable them to do more than note taking during a meeting.
4. We’re seeing recruiters actively use Read to streamline their workflow. On average it takes 700 interactions for a recruiter to fill a single role. By measuring candidate engagement and sentiment across various topics, recruiters can instantly look up what’s most important to a candidate based on their reactions. With candidates, Read creates a better meeting experience by making sure the interviewee has the right interview history.
Tell us about the Read story
Rob Williams, Elliott Waldron, and I (co-founders) have worked together across four different companies (Placed, Snap, Foursquare, and now Read) over the past decade and all of the companies have been focused around measurement and attribution.
With over 500 million people using video conference solutions on a daily basis and yet up to 40% of them are disengaged, the opportunity to return this time back to the people was an opportunity we couldn’t pass up on. AI + Analytics is what will make meetings more effective, resulting in fewer meetings, and better meetings.
Tell us about the technology behind the products
Read takes a multi-modal approach to applying AI and analytics to video conference calls. In real time and post-call, Read converts audio and video signals into engagement and sentiment metrics, as well as applying foundational models around text to automatically generate a meeting summary, list of topics, action items, and key questions.
Using a multi-modal approach, Read isn’t limited to just the speaker, when measuring meeting effectiveness, rather we are able to go broader, but incorporating the reaction of meeting participants to the speaker. By incorporating participant reactions, it is a built in studio audience that helps value the audio and video from the speaker.
Given the type of information that Read collects, how do you ensure data security and privacy?
Read is built on a foundation of user privacy, transparency, and control. With each Zoom meeting where Read measures audio and video, participants opt-in to measurement via a recording notification. If a participant chooses not to be measured, they can decline being recorded. If a user changes their mind, they can type “opt-out” into the chat and the meeting data is immediately deleted.
A privacy first approach to measurement is in our DNA. Rob, Elliott, and I (co-founders) came from Placed, where location data was only accessed with a double opt-in and with a clear value exchange, where we paid users. When Snap acquired Placed, we continued to build on the concept of privacy first, with messaging being ephemeral. When we started Read, we wanted to continue to build on the foundation of user privacy, transparency, and control that we developed at Placed, Snapchat, and Foursquare.
What are your predictions around the future of artificial intelligence, as a whole?
ChatGPT highlights the potential of AI in the mainstream, where day-to-day tasks can be automated or assisted with foundational models. As companies, and specifically startups, build products on top of these models, the story won’t be AI, but automation.
When Microsoft first introduced spellcheck and grammar, it felt like we were cheating as technology guided us to create better content. AI in its current form will do the same.
What would you say are the most interesting business implications of ChatGPT, if any?
ChatGPT has the ability to free up hours a day by automating tasks, enabling individuals to focus on deep work. As an example, Read Meeting Summary automatically converts a video conference call into a summary, list of topics, action items, and key questions, 15 minutes after the call with no human interaction.
While there are 500 million video conference calls a day, do each one of them have meeting notes? No, I’d argue it’s more ephemeral like Snapchat, where we mentally retain select points, but nothing is written, and it disappears into the either. By applying ChatGPT to every video conference call, it makes the meeting more valuable.
What do you enjoy most about serving as a CEO and technology leader?
It’s the ability to take a novel idea and work with a great team to make it reality.
Is there anything else that you would like to share with the CyberTalk.org audience?
Read is available for free to the hundreds of millions of people who use Zoom, Google Meet, Microsoft Team, and Webex on a daily basis. In addition, Read was recently named an Essential App by Zoom and made available to millions of paid Zoom accounts.