Tuesday, 20 August 2019 00:13

Voicea: Meeting intelligence with AI: Interview with Cory Truffiletti, CMO

Written by
Rate this item
(0 votes)

voicea2 5

Imagine.

You’re in an important phone conference with the top managers under your supervision. This meeting will set your company’s priorities over the next six months.  Each manager must present a report. Decisions will be made during the phone meeting.  Conclusions, based upon those decisions and the underlying details must be outlined and prioritized.  Each manager is responsible for a certain corporate function.  Each must be very familiar with how the meeting impacts their own respective department and of course, how everything integrates with the meeting's big picture.  Most importantly, the CEO needs the final report, yesterday.

Impossible. Right?

Not any more.

So, welcome to the world of artificial intelligence married to the latest in collaborative technologies.

Welcome to Voicea.

 

Believe it or not, whether you are a business executive, a politician, a member of the sales team, or a member of the family, you can now record the conversations of each person and virtually seamlessly organize, transcribe, segmentize and communicate the conversation to the conference participants or to others, with instructions, directions, reminders and so much more to make your job easier and more complete.

 And, you can do all of this whether it is a small or large in-person meeting, phone conference, video conference or even town hall meeting. 

You just need a little help from an artificial intelligence digital robot assistant. Her name is Eva. She lives in the cloud.

Recently, I had a chance to get to know Eva test drive the technology and in 21st century parlance, “it blew me away”.  It’s almost like having Google Assistant making calls, join the conference call, take notes, filter what’s important, organizing  future assignments—all in the background,, in real time.  She can expand your information base and finish her job in hours, if not days.  Prior to Eva being on the job, mess less work would have taken much longer to complete.

Hard to believe? Difficult to understand?

If so, I’m at your service.  A few weeks ago, I discussed the Voicea program with its Chief Marketing Officer, Corey Truffiletti, and of course, Eva tagged alongs. Here is the transcription of the interview and the video. 

Hope you enjoy.

Hi, everybody this is  Steve Sabludowsky, publisher of Bayoubuzz.com. And today we're going to be talking technology business. And I'm going to put a little house use this new technology in the world of, say elections and politics with Cory Truffiletti . I hope I said it right. Good. And Cory is going to tell us about the product that he is involved with. In terms of something I have done a 30 day trial, which I think is really amazing. Technology. Voicea so Cory, why don't you tell us about yourself first, and then we'll talk about the voice. I go ahead, please.

Sure. So my background is in online media marketing, advertising, and so on so forth. I've been doing online as a marketing platform since 94. And most recently, I had joined this company to be the chief marketing officer for Voicea. And prior to that, I did a stint with Oracle and a company called Blue Kai, which we sold to Oracle. So I was in the data space. Prior to all that I ran agencies, I was a digital agency person. So it's been a really nice trajectory going in from agencies to consulting, consulting to marketing, and then marketing with data, now marketing with AI.

Okay, great, thank you. So what exactly is working with a?

So for me, it used to look at it from two perspectives. On one side of things, I'm looking at tools that allow me to be more efficient as a marketer. And on the other side of things, we have a product that is AI driven, that I'm helping the masses understand how to use and help them understand how it can be a benefit to them in their daily lives.

YA, I mean, what's what's the deal about AI, if I'm not,

it's a hot topic, obviously, a lot of people talking about it. So AI is really a tool that can help you by augmenting the average day of the knowledge worker or just the general worker in general. And it's really something that can be used to automate certain aspects of your day, I usually used to tell people that there's not enough time today to get all the things that you need to get done, done. And I can automate some of those more mundane elements, some of the things that take your brain about two or three seconds to accomplish, that you could take off your plate, if you can have AI, take those manage those for you, it frees up time to be able to do the other more important things. And one way of an analogy for this is if you think about the Marvel movies, you've got Iron Man, Iron Man has his suit of armor layers over himself. And he's got an AI built into that. And it augments his ability to work, it makes him faster and stronger and better and more access to information. And that's really what AI should do for people to it should allow you to have more access to front quickly, and be able to use that and put that into action a lot more more efficiently.

Yeah, yeah. So at the bottom of the screen, we have low image. And from what I can see, we're actually we have AI in work right now. So So who's this image?

So that's Eva. Eva is our enterprise voice assistant. It's an acronym. And what it does is it joins any kind of conference call, you can also have it on direct phone calls, and you can have it in person with an app that we have. And what Eva does is it joins and analysis itself, then it listens, it transcribes the entire conversation in real time. And it helps capture the most important moments that have been designated by me. And the way that that happens is three different methods. One is that I can use voice commands, which is very similar to having an Alexa in a meeting so I can say, okay, Eva, the action here is for me to follow up with Stephen and do something that we haven't done before. Thanks, Eva. So that marks that moment action item, as an action item in our highlight reel. I can also do things like say, okay, Eva, remind me to call Stephen again later today. Thanks, Eva. schedule, that's going to go into my calendar, it's going to schedule reminder for me, and it's going to send me an email to remind me about it later on as well. So there's a lot of those types of actions that I can do using a voice command. Now the second thing is that I set up what I call trigger words and events, which are passive words. And these are just terms that I typically will use in a meeting. And even will highlight those and captured as as a moment. So for example, one of my trigger words is we should, so I'm in the meeting with my team. And I'll say we should create the presentation to send through on next Friday, that's going to get marked as a moment in my highlight reel and captured and extracted as a point of information. And then the last thing is that there's a UI that I can access on the web, which allows me to push a button. So if I'm talking or you're talking, what you're saying is the most important thing ever been said before, I just push that button, you can see that Eve is now taking an action and recording that moment in real time. And then when it's done, I push the button. And that is now part of my highlight reel. And I end up with this full recording of the audio, a full recording of the video, I end up with a full, very highly accurate transcript of the conversation. And I end up with this highlight reel that has the most important moments captured and highlighted.

Yes, so at the very bottom, you can see, the audience can see that basically everything I'm saying right now is accurately, I must say, accurately being recorded, maybe I would say I've been doing voice recognition. I gotta tell you since 1990, when it was like 20 words a minute 90% error.

Uh Hmm.

And, you know, this is this is accurate. This is accurate. And so we so basically what's happening is it is it is capturing our conversation. And it is placing it in a way so that later on I can go back. If I started this conversation, this is a conversation between the two of us and it could be a conversation among 10 20, 30 people.

Cory 

Yeah, yeah.

And so I can go back and look to see exactly what the key words are. trigger words are, and be able to use that in some way. So you're going to maybe use a little tour.

Yeah, sure. Let me share my screen with you if you're not right now.  See here. Right now you should be seeing my my Google Chrome. Is that correct?

Yeah.

So you're seeing right now I believe. The Voicea, the platform, correct?

Yes.

Okay. So on the left hand side here, you have what we call the meeting captions, this is the real time transcript of what's going on. When I mentioned a minute ago, there was a button, I could click the tab. That's this button right here. Now, that is something that's getting captured and put into my highlight reel. And then down here, I can also add notes. And those will get dropped into the regular conversation that we have as well. So this is what it looks like during a meeting where you have the meeting highlights being captured here on the right hand side. And you can see, you know, this one was captured as a note. This one was an action item created from a voice command. So it also denotes how these meeting highlights were being captured in the conversation.

Okay, so right now, let me just make sure that I fully understand. It says meeting captured to the column to the my left meeting captions.

Yeah.

Okay. So that's what you refer to in meeting highlights. Once you tell us the difference between the two.

Yeah, so the captions is the real time transcript of the conversation that's going right now, on the right hand side, or the notes they're getting captured. So once again, if I say, okay, Eva, the purpose of this conversation is to really show off a demo for the folks who are watching Bayoubuzz. Thanks, Eva. You see how that got captured on the left hand side. Now it's going to pop over here on the right side as a highlight. So there it is, the AI captured that just from the conversation. And then it also highlighted that it was because the word action item was used in the conversation. So that's where the meeting highlights go during a meeting. Let me show you what it looks like after meeting. I think you'll find some some

Sure. Sure.

Well, after the meeting, you go to this past meetings tab, and I have one open. So this is an example of the past meetings page. So when you click on any of those past meetings that showed up, you're going to come into this, and this is the first thing that you see. So here is the word cloud, the word cloud is an overview visually of what the conversation was about. And if you click on any of these terms, it brings you to that moment in the conversation

Recorded voice 

is August or July roughly flat

That's what we're talking about audience or if I click here,

Recorded voice

our web.

So this gives you a quick overview and ability to jump to any part of the conversation, you can see here the running transcript that's associated with that. So that's the first view of past meetings page, you also have here, this is the full audio. And if I click here, I have the full video of that conversation as well.

Plus, the plus you can download the transcript. If that's what you're looking for you

Yeah

to use a transcript. What I did, quite frankly, is, is I did a show and took the transcript, and basically just put the transcript up on as content.

Yeah.

And I mean, it's that fast. It's actually faster, I found it the platform, it's a Google Platform?

NoNo, this is all entirely built by our team. So what's going on there. So in the meeting, in real time, there's a single transcription engine an ASR that we built, that takes the conversation, transcribes it, after a meeting, we have what we call an ensemble processing, which is we also have four or five other ASR'S , but can be can be used to review and transcribe the conversation. So what happens is that the meeting goes through five, automated speech recognition platforms. And then the ensemble picks and chooses the best and most accurate elements of each one and stitches them together. And so the reason we have so many is that some are trained based on audio quality. Some are trained on accents, some are trained on different types of vocabulary. So it's stitches together, and you end up with something, it's about 90 to 95% accurate for real world meeting content, that's based on the different types of audio and everything else. So what you get afterwards, about a half an hour after the meeting is a fully reviewed and process version of the transcript. And as you pointed out, you can download the audio, you can download different versions of the transcript. And for the people who want a 100% accurate version, you click this button right here, and you can go get a human transcript, there's an additional cost, because we run that through a partner with the versions that we have, is all completely machine driven. But if you want a 100% accurate, one, human will do it as well.

So if you want to go back, let's say there's 10 people in this conversation. So will it automatically pick up who saying what? I see here in terms of past? It has different email addresses? I believe.

So

will it will do that automatically? Or do you go back and say, okay, Joe said this, you know, Tom said that, go ahead.

Yeah, so that's called speaker tagging. So the very first time you're using our platform, what's going to happen is that you'll come into a meeting, and it'll ask you to do a bit of speaker tagging on your head. So what you do is you would scroll through here, and you'd click on this name here, and you would attach it to the person who it was. And the first time that you tag somebody in our engine, they're now going to be voiceprinted. And every time we see them in the future, we'll be able to recognize them and associate their name with that email address or whatever name you've given to them. So the first time that you do it, it then carries through in perpetuity for the rest of the users of the platform. But it does require that very first time that you have to tag them.

So that's the artificial intelligence it learns from the first time. That's, that's incredible.

Yeah,

it really is incredible. So let's, let's talk if you don't mind me, I, again, I. I used it, I thought it was just absolutely really super-- I'm looking at the different subscriptions now, at this point in time, and I'm trying to figure out exactly, you know, how can I apply this technology, the different tools that I can use with it, like zoom being one of the interfaces as right now. And so I'm going through that process, in terms of how can I apply it, what type of environments. One thing I did, in my example was, I was sitting at a meeting with somebody, I took out my my app with their permission, and we just recorded the conversation, then afterwards, I went ahead, and I shared it with him.. Second, secondly, in a telephone conversation, I told them that we're going to record this conversation, and give them a copy of transcript. got permission, of course, and, and, and then as soon as the video is over, I mean, I was very surprised how quickly it took, you know, was able to go and share that information, the conversation transcript, or with the video of the conversation and, and then be able to do whatever we want with that. So those are just, you know, a telephone conversation in a very real time one to one, say, sales conversation was just say your meeting that I might have with with somebody, what are some of the other applications and if you don't mind, I'm going to try to translate that, transfer that into like a more of a political environment.

Okay, so the the best use cases for us typically have been around internal and external meetings and companies. So if you're doing internal to product management, project management types of meetings, sales, people use it a lot on sales, many marketing teams, also executive level conversations, anybody who wants to be able to utilize this, this information and the discussion and turn it into content for an extended team. Those are the cases and so you started talking about something, I call it our talk matrix in our action matrix. So from a talk perspective, anywhere where you're having a conversation, so if you're doing zoom, you're doing Skype, you're doing Google meets, you're doing Cisco. With WebEx, you're doing your freedom conference calls and dial pad, whatever platform you work in, as well as direct phone calls and in person, because you can use our mobile app that enables you to take anything from talk and turn it into content. Now from an action perspective, and I'm going to show you this second page of the UI as well can I get some partially important to that. But you can see here that when the note is taken, and you can edit your notes, and you can drag them here. So you end up with a highlight reel that is specific to a team. When we have these things here. These are all the integrations. So the integrations allow you to push this as content to any platform where your team works. And that's what we call the action matrix. So for example, you and I jump off a call, if there's follow ups on this call, I'll just click this and send you an email with the follow ups and a direct link to the content inside of VOICEA, if my team operates on Slack, I can just simply push the button for slack. And it'll go create a an update in a channel on slack. Maybe I want Salesforce to be updated, I can go update Salesforce Trello, Asana, JIRA, there's about 40 different integrated platforms that we have. Because for us, we think it's more important for you to be able to get the meeting content to where your team does the work. So for us, it's not about you coming into our platform, it's about making sure that that conversation can be converted into content, and put into action as quickly as possible. That's really that's the whole goal of the company is to make it so that anybody in a meeting can be focused on the conversation, not have to ask, What was that again? Or can you repeat that, because you have this virtual record of the conversation. But for the people that were necessary for the follow up, and the all the things that have to happen after a meeting, you can decrease the amount of time required for all that follow up, because it's now a tangible piece of content. It's not this ephemeral, temporal moment in time that you really wish yourself “Oh, I wish, wish Nicky was here. And I wish john was here. So they could have heard that section. So they could have followed up on it properly”.

Wow, I mean, seriously, as a big wow, because the thing is that in terms of applications that I can think of that I haven't used thus far, would be like, you know, having a, say, a town hall meeting, example. And let's say it again, I'm going to transfer this now to the political space. Because, you know, a lot of our content is related to that. And so, a town hall meeting, once you tell us how it could be using a town hall meeting.

So I mean, we I will caveat with we don't do a lot right now, right? government by any means. But I think for a town hall conversation, what would be great is to be able to have a record of that conversation. So you can imagine a situation where you've got our app on the podium, and I have done this for the app is on the podium, and it's recording my conversation. And then if you assume that anybody else in the room is going to be handed a microphone, to be able to ask questions and to engage, and you can take the audio directly off the soundboard. Or you can take the audio that's record off the soundboard and uploaded into our platform, which you can do, then you end up with virtual record of the conversation, you end up with content that can be accessed and it can be shared, maybe with people in that in that district. You want to have access to information for later. You know, I think that's something that potentially could be done. Sure.

Oh, absolutely. And of course, a lot of town meetings now are digital, I know that I've done, I can't tell you how many hundreds now of different say, conversations with 2,3,4 people and using that as content or whatever. That's a great place. A great, great way to use this. So you can have it. You can use zoom for an example and have a digital town hall meeting.

Yep,

able to use that content. Why don't you tell us more about that possibility?

Well, I think it's the same as what I was talking about even in an enterprise environment, is that a meeting or a town hall? It's one of those things where if you weren't there? Did it make a sound right? if a tree falls in the forest, did it make a sound, there's nobody there to hear it kind of thing. If you wanted to attend that town hall, but you couldn't, you still want to be able to know it was talking about you still want to understand the context. Because it's not always purely about the decision of the actual comes out of a conversation. Sometimes you have to understand the context. And I think what this does is it preserves that context, in any kind of meeting or conversational environment, and enables you to understand exactly what was talked about and how. So rather than just walking out and maybe taking something out of context, not fully understanding what someone meant by a statement that they made, you now are able to do so and it's there. And on one side of things I imagine you could probably use to hold people accountable. But on the other side of things, I'm hoping that it would allow people to have all the information because I think, too often, you know, if your, if your content primarily is around politics, which I think too often we rush to judgment, because we don't understand the context of what the conversation.

Sure, in terms of promotions for what I'm thinking is this that, let's say I have a, let's say I have a town hall meeting and town meeting last about, say, half hour. And as this this meeting is beginning to do, you know, basically, if part of the town hall meeting dealt with let's just say--say health care, if I use a target word, I could get all the content related to healthcare, able to transfer that information immediately to those people that are voting, recognize, have an interest with health in health care. Likewise, if the conversation was dealt with health care, and or say some other subject, let's say Homeland Security, same thing, I can go and do a deep search, find the information related to Homeland Security, and within seconds, be able to funnel that information to those people who just have an interest in Homeland Security. That is an extraordinary, quick way of getting information, not just the text, and transcript. But now questions that are like you said contextual questions that are asked, and the whole conversation related to that particular category?

Yeah, yeah, I think there's a lot of applications of this, this kind of technology across the board. So Sure, absolutely.

Absolutely. Is there anything that say, we haven't talked about that you think that the business audience of local audience, just the people involved in, in trying to understand technology and media, you know, is anything that we haven't touched upon?

I don't know, there's anything we haven't specifically touched on, I just think it's interesting, because what you're doing is you're you're raising a new use case, a new opportunity for how our product can be used. And this is one of the things that we are constantly talking about on a daily basis. Because while our product itself is probably pretty complex, there's a lot of elements to it. And for us, it's about helping people understand just that you can try it out that you can use it and you can get an overview of a conversation. But there's so many different myriad of ways that it can be used, that it just comes down to how do we simplify the understanding of it. And you know, it's a really at its core, it's simple, it creates content, from a conversation that is now searchable, affordable, accessible, etc. And I think we're just we're trying to continue to teach people how to do that.

Sure. So it's basically agnostic and its technology approach. And it's just a question in any type of environment, any type of industry, any type of institution can use it.

Yeah, we have customers that are using us in field construction. They go out there and they record their notes for themselves, and then they can send it to their team. We have people in a corporate enterprise environment, we've got people in coaching consultants, I know there's real estate agents that use it. So it's pretty much all over the place.

Yeah. Again, I had a full month of opportunities. Take a look. And I really am glad we had this conversation. Because I can see, you know, a quantum amount of new opportunities that I hadn't even thought about.

Yeah, absolutely. And as you come up with more those ideas, Stephen, please feel free to send us it on inspiration.

Absolutely. Absolutely.

Last modified on Tuesday, 20 August 2019 21:22
Login to post comments

Our Past Webinars

allee 5


bern jim3 2


buisson elections4 1

 

 

cherv james williams 3

Dead Pelican

Optimized-DeadPelican2 1 1