top of page

AI Voice in the Workspace


Awesome. So, before I lived about 20 years in New York and I moved about four years ago to San Francisco and I literally live like half a block from here awesome views of the bridge. So, I love the marina cool. So, first I want to get started by asking if anybody knows what the following image. is anybody want to give it a shot. okay, this is the cover of a 1960s report on speech recognition by the IBM TJ Watson center. the reason why I bring this up just to give you a little perspective. it's been 60 years since engineers’ scientists have been working on getting to that point where a bodkin has a natural conversation with a human now in the 60s, they got you know, they made some strides but they didn't really get far enough. So, I’m going to take you now to 1999 during my summer internship. I think I’m a freshman in college and I did it at the IBM TJ Watson center in Yorktown heights. I remember two things out of that experience one extremely memorable. I remember my first day at this internship. I walked into the cafeteria and I just see a sea of just white hair as far as you can see it looked to me like I was probably the only person under the age of like atm that's not true. that's what you see when you're 19 the second piece, which is really exciting. Let me go to the next one. you're going to I got to work on the text to speech team of doctors and a ham. oh, he's considered the father of speech recognition globally until recently. he was the global CTO for IBM on speech recognition. now that was 1999. I learned that about 2006 the technology wasn't quite there to get to the moment where bod can have a natural conversation with a human and some of these themes were paused. So, I’m going to take you now to hear anyone wants to take a guess who these guys are Adamdag and I think gruber I don't know how you pronounce it. anyone was a give it a shot. all right, half of the audience probably hear more and have an iPhone. these guys are the inventors of Siri 2008. they started Siri 2010 that they get acquired by apple. I’m not going to lie. I can tell a little bit of a story because I loved it initially. this is this is revolutionary is a magic moment. it's the first consumer hit application end-to-end voice. So, you speak to Siri and it gives you the answer by speaking to you. the coolest part of one of the coolest part of the story is that actually Siri initially was meant to be you speak to Siri and it will give you the information on your screen and they decided to delay the launch to make it magical and eventually they made a voice the voice. oh, it's that part. I really love they plugged into like nuance, which is like a 20-year-old company in Texas speech or more as well a little nerdy fact that I love about that story. So, huge magic first time. you see an end-to-end voice magical application on everyone's phone. let's go to the next one. but of course, there is some limitations. you have to know. how do you ask Siri certain things and the bots still doesn't really quite sound perfectly human? but it so this anyone Woodson boys or anyone's in chat bots or any but since in Silicon Valley pretty much remember this moment 2018 google announces duplex. They do what demo where a voice pot has a perfect conversation with a human setting up our hair appointment. it's the really the first moment where you see a natural conversation the person on the other line, that's not really believe that they're speaking to a bot. they think they're speaking to a human. this is the first moment. where the this is what I guess you can call it the sputnik moment. this is where for first time the level of technology mailing you with speech to text to speech gets to the point where sophisticated enough that you can actually start building serious business applications with ai voice and that's actually why I’m here. I’m going to share with you how AI voice will dramatically change the way business leaders work. my name is Ricardo Garcia Maya. I am the founder of VOIQ. we help sales and marketing teams leverage voice bots to create warm and relevant conversations over the phone channel with their customers and prospects. and I’m going to start sharing my favorite top 5 ai voice innovations for business. the first one is the voice dashboard. So right now if you're a CEO if you're a sales leader, if you're a leader in the organization, we have the visual dashboard, which is incredible puts the information or a fingertips now even better than a visual dashboard is a voice - were so I’ll tell you a little bit about what you can do. So, you will soon enough be able to get information about your business whether you're jogging whether you're driving whether you taking a shower and currently let's say that you are driving commuting early in the morning to work and you are just thinking about you know, what I would like Diane’s numbers and sales from yesterday, right? currently, it involves pulling over pulling up your phone pulling up your laptop going into a website go into the dashboard internet query pulling that information right now with a voice dashboard. there we go. you should be able to just speak to the voice interface. in this case in your car could be a watch it could be your shower and you can just ask how we're of Nancy’s numbers in sales yesterday, right and information will be provided to you. So, in terms of in terms of efficiency is going to be a game changing. the reason why I bring this up I bring back I mentioned dr. animo earlier. he now is working a company called praying and they call it a mental intelligence. but this is what it is. you'll be able to ask a voice assistant for any type of information and pull it up regarding your company. second is the voice meeting assistant. this is huge specifically for me. I suck at taking notes in a meeting. I’m either I can't find it extremely hard to be. I sent in a meeting and at the same time capture or who are the stakeholders. what are the tasks? who's going to take on those tasks? who are did we did we measure success in this meeting was this successful? so these are the typical kind of scribbles that I do and after the meeting I’m like what the hell did I just that mean? I don't even know what I write. So, for me in for about millions of Americans are about 55 million of meetings every day in the united states. So, this is going to be game changing their several technologies that listen to your meeting. I baked a gate give you a report after the meeting in multiple channels. So, on the first one, there are several companies that are transcribed listening to your call conversations call meeting so many of you are familiar with the company chorus. then you also have the same on video meetings you have Kong and now you have face-to-face meetings and you have applications. but listen to these conversations and give you the results after the fact, so that's one of my favorite ones. third is search navigation. you probably heard or been exposed to this one by maybe adobe the example that I like here in terms of collaboration. if you are let's say a designer and you're working with a couple of other designers in brainstorming to some creative idea and you're browsing through a catalog of you know, stock photography right now. It involves one designer tied to a keyboard and a mouse and everybody else kind of yelling at them with ideas right in every search query it. It's you. Have you had to start pretty much from scratch, right? text and you get these results one by one now with voice search and navigation. you'll have you'll be able to bring more collaboration to this process. So, you'll have several designers in a room and you won't have to have one single one tied to it. you can if each one of them can collaborate and share ideas directly to your interface. this is already in place. you probably have seen adobe has this feature. So, that's one example of how AI voice will dramatically change. is that Voice Authentication is using your voice as your password? This is great. I mean the average professional has about 200 passwords to deal with of course. we have plenty of password managers that try to help make our life easier. I just hate when I actually forget the password to my password managers have been a couple of times and you kind of have like a mini heart attack when you when you're able to leverage your voice as a password, I mean they easy examples to think like oh, you know, I no longer need to enter a password to a website but it goes beyond that. I mean, I’m seeing it from a workflow perspective. I don't know how many of you participate in if you're a midsize company or enterprise company and you want to jump into a video call. you have to find a conference room. you have to login your credentials. you have to go into account and find the video application find the conference number and there's that's why so many people are usually late to this meetings with voices education allows you to just walk into a room and just say patch me into my 10:00 a.m. meeting, right and it logs you in in multiple platforms and get you there. So, huge game changer now. I’m going to get to my favorite. obviously this is what we do. we're experts in this area voicebots for business calls. So, the easiest one someone mentioned here earlier the one that comes to mind is its ivrs right? so if you have if you call a vendor you call your bank your doc in these ivrs that don't quite understand what you're saying that hang up when you throw you in loops is it's usually a nightmare now soon enough you'll be able to have a ai voice bot that listens to basically on the first ring. whenever you call the vendor on the first ring, it will answer and say hi. how can I help you? and they actually will that thing will actually be able to help you and answer all your questions. now this technology is not here yet. ever you are able to use the same voice bot to do something with a little more of a narrow focus such like marketing experts sales manager sales automation can use voice spots to make sales and marketing calls and that's currently the use case for voice box right now. And that's what we do. So, the reason why it is the technology for ivrs, it's not here. it's because the scope is so wide when you are just asking somebody, how can I help you? right. however, I mean you can have somebody calling, you know an iv ours as well. you know, my grandma was here and she brought the cat spills the water my modem and now it doesn't work done. right the nlu is not going to understand that it will maybe in 10 15 years, right? so for now though, if you narrow the focus of this of this conversation, then you can actually use this voice button and that's the way that we use it when you use it for marketing and sales calls the voice but is I wrecked think the conversation right so it's asking you. hi john. thank you for attending Ricardo’s talk. I noticed that you register for timorous workshop. will you be attending right? that's a very narrow focus and you will the nl you will be able to really understand it be prepared to get what you're saying. So, that's the big difference between those two. I’m going to take you in a little bit deep dive in this one because this is our expertise and I’m going to share with you what to expect from a at form that allows you to launch voice spots for sales and marketing calls. So, the first thing it's that it's a sas platform. it's online you go to you go to like you would to go to any other sort of sales or marketing automation platform to create your campaigns very similar or two. if you're looking for a chatbot automation platform you have drift if you're looking for an email information. You have outreach you have MailChimp you have active campaign, right? and if you're looking for a call voice spot on. should platform. well that's us where the first ask platform that allows you to do that what to expect from this platform first that it's human sounding. So, out of the box there their tents of providers of synthetic voices out of the box is not going to sound perfectly human, but it's going to sound human like you can make it sound perfectly human with a little bit of time you should take for that demo where google took about four months and several engineers just to make it sound. perfectly human to just set up a hair appointment. right? If you're having a one or two-minute conversation, you can make it definitely sound perfectly human, but it's going to take some time not months hours. it's going to be conversational. So, again, the narrow scope allows you to be extremely relevant eyes is Linda great, you know hide this is Robert blah blah blah blah blah, right. I will you be are we still on for the meeting at 3 p.m. then it's integrated to crm which makes it even more relevant. So, it's pulling information from your crm which in this case is more like a real estate type of use case. hi, josh. just want to make sure that we're still you're still coming to the open hands us for so it's confirming stuff for a real estate agent how transfer you can actually transfer the school to a human. So, if if you are in this case, it's somebody that join a group on Facebook on to prospectively become a franchisee of a gym.

So, the bot automatically is calling somebody who joined that group and it's saying hey, thank you. would you like to learn more about do you have time right now? I can introduce you to an expert that can tell you more, right? so it's more lead generation prospecting the next one voice cloning. this is pretty cool this soon enough. you'll see brands that have a strong voice kind of like the gecko in the geico commercial will be able to call you after you fill out a culture. to action on your side or when you know, when you are approved and say, you know, I Josh congratulations mate you approve, right? so that's what's coming down. the pipeline in terms of voice cloning workflows triggers. This is one of the coolest things for me and that is that voice pots for calling are not needed to be triggered one by one or in groups by a human but they are triggered based on actions in based on what you're saying that call it triggers other actions as well. So, this case this was triggered by carrying downloading an e-book. right? so avoid spot automatically goes in culture and ask her, you know, thanks her for downloading an e-book and you know, ask her if she's interested in learning more. simp simple objecting handling, right? oh, you know, I don't know I learn about your product. but you know, it's a little bit too expensive. Well, we have installment plans. so, you know, if you want again, they're not going to take close the sale of voice about will not close the sale avoid spot will not take the credit card, but they will patch you they will transfer to somebody to be able to close the deal so you can definitely train your voice bot to handle objections. and of course, like any other sas platform you have your dashboard that tells you what works what doesn't and you continue to improve the learning of that campaign that you're launching? and way more things that you can do one of my favorite ones here is they be testing so before if you wanted to know what's the best type of profile of a call agent calling a prospect or prospect or customer? You know, would it be a woman between the ages of 35 to 45 with a Texan accent or a man between the ages of 21 to 25 with a neutral accent is used to take you about a month testing multiple sort of personality and profiles now within hours you can test the perfect. can in the perfect profile for the type of person that you want to be contacting a prospect or a customer selling or prospecting for a specific product? okay, and the use cases I mentioned the sales and marketing customers are says pretty much this is it's a brand-new channel. this is the first time in the history of marketing that for the first time in the history of marketing. they have access in control to manipulate the cold channel, right? so marketing automation experts have also workflows of emails text chat bots and now they have a whole new channel. So, they're having a lot of fun leveraging the cold chatter for voice pots this I’m to give you one use case. this is new to everyone including market animation cells animation as of a year ago. you couldn't do this right technology wasn't there for you to do this? this is a typical example from marketing type of customer pretty much any business in the world that has a call to action online in this specific scenario. It's a car insurance and its they get about half a million to a million visitors visit out entering the call to action and they're not going to have humans in this early calling. you know, if you're calling if it's a million people filling out your forms, you're probably making four million calls and it's extremely expensive you have humans, but you can use voice bots and the voice box will call at the right time a hundred percent of everyone that needs to be called a needs to be prospected. So, in this case it says hello. hi is this Karen? yep. hi Karen. thank you for a person to learn about your car insurance for your Toyota corolla. right? we know that because they feel the form they just did and they have one additional qualification question does your car? have more than a hundred and fifty thousand miles in this case, you know, she said yes or no or whatever it is and depending on the question. that's a great you qualify. I would love to do you have time right now? I can introduce you to an expert who can get you set up with your care insurance, right? let’s share you like a typical use case from a marketing perspective now in just two years 30% of enterprises will be leveraging voice pots for sales marketing customer success customer service. So, I hope this this inspires you and start being an early adopter of voice bots for sales marketing customer service any type of stuff happy to answer your questions. thank you. thank you very Carlo. three questions thanks Ricardo. I’m really looking forward to a future where this is. this is the thing you call in to a big company and you get a got artificial intelligence helping you through my question and it might be a question for a lot of people in this room is with regulation and possible disclosures down the road. what have you seen so far and where do you think that's going to go? yeah, so actually the regulation that's coming into place. We love it, which is the need for businesses to disclose or register a certified. their name so instead of you getting a 1-800 number is going to say the name of your company. and the reason why you're calling that's extremely useful to us because our customers 99.9% of the campaign's that are run our platform are triggered by somebody already interacting with your product. right and they are they want to talk to you as a prospector as a customer at that time and they're competing with unsolicited calls with 800 numbers, right? so you might be you just feel that a, you know, a form online your think you're in that moment. you actually would welcome to learn more right? but if it doesn't say who was calling you then you could think that it's not relevant to what you just did. right? so in terms of regulation, we can't wait to be able to have the next time when you have a caller id. it should say the name of the company and the reason for it will you know already it's releasing in beta the original kind of technology also.

I’m sorry google. I don't know if they mention it earlier, but they mention it in in the chat bot conference in New York. but yeah, it does that answer your question. So, the question is well, so the question is do everything that we do is opt-in or and okay, that's a great question. those are two important questions. do you need to disclose you're not human so when you're calling somebody and there is a reason as to why there's opt-in. you can call them whatever you want right you what do you need to disclose it? whether the call is recorded. you don't have to disclose whether the call. it's a voice pot or a human today. that's correct, and I can get into a longer story about like robocalls and the difference between like a robot just to put it out there. the robocall is a call is a pre-recorded call. it's universal and is sent to all of you. So, just going to say hi. you need car insurance. I would have no life insurance, right? so not customized to you is it's a whole different thing. So, regulation that you've seen over the past like 40 years is based on those type of recordings that are completely unsolicited. I compare it to email and spam any respectable company in the world knows that is not a good idea. if you ask a sales or marketing manager to just send 2 million emails to people that are not in your icp your ideal customer profile because your brand hurts so the same thing applies to phones right if you're a savvy marketing sales expert you will not call people that you should not be calling that are not an icp that are not that they're not warm that they're not expecting that going that it doesn't provide added value. So, same rules it just that most people are familiar with kind of unsolicited calls the way the people we're very familiar with spam like 25 years ago before he kind of got under control. anybody had like a viagra 10 emails about viagra and you know your aol account 25 years ago. and now that's not doesn't seem to be a thing filters technology kind of solve that so yeah, I don't know if that answered the question, okay. any other questions, why is the mic is so you're so a solution how the inducer would be able to access your voice part. do they need to have any kind of 1-800 number or download any app you can you repeat the question is can you hold the mic so are actually I would like to ask like kandak your solution can be integrated with the existing iv our infrastructure? how can businesses leverage this? yeah how inducer can access your voice? yes assistant. okay, so we need to have any specific dial-in number or they need to download a mobile app. So, business or the possible ways. this is user platform in the same way that you would use a sassy male automation platform. they log in they create campaigns and it's mostly I mean, all of it is outbound. So, this is very different from ivrs that you'll only look more like a call back. sorry. is it more like a call back during that time? we are making the cargo ship fluid. sorry, I didn't get that. can you can you ask it in a different way? so if it is outbound means actually. Your intelligence system is making a call to customer and provide a fluid assistance. is that the case? yes, so it's tied to see a ram’s type to salesforce has tied to close that io is tied to any of those sales and marketing automation platforms. it's tied to we're natively integrate to hotspot. that's a great point. so, yes, you can drive the workflow inside the software that your ready work with if you want a sophisticated can paint and you have to login into our platform and really, you know hone in and train it. but you can deploy it and create a workflow from wherever of those platforms that you already used for sales and marketing automation and I’ll give you an example we are right now onboarding I guess that the biggest market place in the world and initially the use case was for marketing and sales, but as they learn the etiquette that the capability of leveraging voice bots they are now are looking to insert it first in vendor. I meant so if you have a team of thousands of you know, home services plumbers around the united states and you want to keep them updated in keeping the schedule. It's an additional channel in more effective channel than text and email. So, none of these this is just an additional channel to email text and you need to understand from a sales and marketing if you're a sales and marketing expert you every channel there's a specific way or use case in your company for it. right? not every company could benefit from tax. another company benefits from calling it works for many companies who work for different, you know thousands of different use cases. We're just learning of all the thousands of use cases and we learning from our own customers, right we say this is great for sales and marketing they said yeah, no for sure. let's do that. but can we also do x and y right so we're learning a lot from them and it's kind of like I would imagine people asking about email like 20-30 years ago, right? like, oh I know that I can use my email to like talk to my grandma and you know, and we are you know this is email for like sales and marketing? they're like can we use email for customer success? and we're like, yeah and they're like, can you give you an example, right? so it's a channel we started with sales and marketing but it's a channel that you can use this for any department. thank you. So, you spoke about the voice being straight out of the box. this is on the far end. But what about on the on the front end, which is the speech-to-text itself like a noisy. environments and even if this are one other person talking, how does the can you speak a little bit more about your platform? and are you addressing that problem because it's a significant user experience problem. yeah. So, right now is super early for I mean, the reason why google was able to do that duplex demo is not because we're I mean, they're amazing, but they didn't do the legwork the technology in a I have helped every single one of those area text to speech to text and will you write and they're all coming together. you couldn't do what we do. If it wasn't for those the advancement of ai applying to those three different areas text-to-speech beast attacks. and will you and in addition to that the telecommunications that having company like twilio right that before you have to go through huge telcos to build an enterprise solution. now, if you're an engineer, you can use twilio. you can use any number of other signal wire to build a telephony application for the enterprise. So, all of these technologies came together and we right now are super early. one of the coolest things about why I like Siri is because there were tying in initially, they were focused a lot on the nlu, but they were leveraging third party for any number of things right and that's what's happening. it's super early right now. it's super fragmented you're starting to see silos of companies that that only focus on cloning voices, then you're focusing. There's a whole cycle of companies that only the planning to be billion-dollar companies in the only thing they want to do is build synthetic voices. there's one that only do text-to-speech. the other one does speech to text and every single one of these silas already becomes potential billion-dollar companies just doing one thing right so there's many players and we're leveraging a ton of these in addition. obviously the sas platform we have about 3 million calls from professional sales agents to to train our voice pot. So, we have some you know some experience in that space and the sas tells you how to improve the voice buzzing cetera, but we're definitely leveraging a ton of technologies from the top leading providers, you know of the world in each of those areas and we pick them we picked the best and we're able to switch when somebody does outdo the other one we switch immediately. So, that's a good question. thank you very much round of applause when I got it off.



bottom of page