The following interview is a conversation we had with Ashish Nagar, CEO of Level AI, on our podcast Category Visionaries. You can view the full episode here: $35 Million Raised to Power Contact Centers of the Future

Ashish Nagar
Thanks, Brett. Nice to talk to you as well. 

Brett
Yeah. So before we begin talking about what you’re building, let’s start with a quick summary of who you are and a bit more about your background. 

Ashish Nagar
Yeah, so I started the company about three years ago, and before starting Level AI, I was a product manager and Amazon’s Alexa team. Did a bunch of things there, including spearheading some of the early conversational AI efforts. And then prior to that, I did a search startup as an early employee and have been doing technology startups for a few years in the Valley, have background in business and engineering, did a master’s degree from Stanford and business and engineering. So have always worked on the intersection of technology and business, and that’s what gets me excited about working on. Lovely I. 

Brett
So if I have any issues with my Alexa, is it fair to say that exactly? 

Ashish Nagar
Yeah, exactly. I can also moonlight on Google Assistant TV. 

Brett
All right, well, I’ll be texting you next time I have an issue question here. Perfect to start off to, another fun question, I think, would be what’s the state of conversational AI. And what’s the state of conversational intelligence today? How would you kind of summarize how things are and how has it evolved over the last five or ten years? 

Ashish Nagar
Man, that’s a big question, but I would say overall, it’s a super exciting time to be in this space. At this point, there are new research ideas and new models, large models being put out every week or every few days, if not every hour. So we are almost in the golden stage of conversational AI. As we speak. And this renaissance almost started, I would say, three to four years ago, maybe five years ago. But in the last two or three years ago with these large language models, things have rapidly changed. I feel like still, there’s a lot to be done, even with all the momentum and new developments and the noise. Sometimes over the next ten years, there are some really hard problems for machines to solve. For example, for example, being able to read and understand fiction, being able to come up with genuine intelligence, being able to understand long dialogue. 

Ashish Nagar
All of these are hard problems which are still people are working on. But I’m really optimistic over the next ten years, things, important things, will change here. 

Brett
Nice. That’s exciting. And a couple of questions that we like to ask just to better understand what makes you tick as a founder. Is there a CEO that you’re studying the most now? And if so, who is it? And I’m telling Founders now. It can’t be. Elon Musk. 

Ashish Nagar
Yeah, that’s a good one. So my answer is equally boring, but I think important. So I worked with Amazon for a couple of years, and I was strongly influenced by how Jeff Bezos runs the company and has been as a founder for the last more than two decades. I could talk a lot about what I’ve learned from him, but those two years at Amazon was one of my best business education and entrepreneurial education. 

Brett
And if you had to boil it down to maybe a few lessons that you learned from your time there, what would you say those are? 

Ashish Nagar
So, a couple of things. One is there’s a lot about bias for action is one of the key values at Amazon. And that’s essentially this idea of planning, but measuring all teams and all individuals by the impact they create and the speed at which they create that impact. And it is easier said than done. But Amazon embodies that value in everything they do, and you see that from the multitude of things which they launch. So that’s one of bias for action. The other is the power of small teams and small amount of investment group big things. So there’s this famous two pizza team about Amazon, but I kid you not like my small team of 1520 people when I initially joined, was servicing a quarter of all of Alexa’s conversations around whether local search and so on. And it’s just incredible to see, and especially in this time, when tech layoffs and loaded organizations are so much being talked about. 

Ashish Nagar
Amazon is a great counterexample at every scale. So you go in and see how a small number of people are doing really big things. And I would say the third thing, if I can say that they talk about customer centricity. But how deeply embedded is customer centricity in their value system? So an SVP or a product manager or the lowest level employee can question any decision based on customer centricity. Are we doing the right thing for the customer? And again, it’s easier said than done. Often with politics and not looking to step on each other’s toes, people don’t say the right things. But Amazon, if it’s customer centric, you should not be scared. And those three things are really critical as an entrepreneur. Like doing more with less, bias for action and putting the customer first. 

Brett
Nice. I love that. 

Ashish Nagar
Yeah. 

Brett
I feel like so many companies talk about being customer first, but as the actual customer, you can rarely feel that. With Amazon, it’s the opposite. You can feel that’s not just some core value that’s hidden somewhere. Every time I interact with Amazon, you can tell they really care and that’s something that they’ve really put into action. 

Ashish Nagar
Yeah. One other thing about Amazon, which Jeff talks about this a lot, and there’s a lot of talk about AI is changing rapidly. A lot of changes happening in the space, in lots of spaces. Right. But one thing which Jeff talks about is ask yourself in your particular space what will not change in the next ten years? Right. Or the next five years? And if you ask that, the business strategy becomes really simple. So for example, for Amazon, it’s like what will not change is prices. People will always want low prices. People will always wanted to get things delivered fast, no matter what the technology, underlying technology is. Right. And if you ask that question, as any startup founder or employee or product manager, it really clarifies your areas of investment. Where should I invest in things which will always be there as part of my business and where I need to get a competitive one? 

Brett
Nice. I love that. And what about books? Is there a specific book that you could say has really influenced how you think as a founder and as a leader? 

Ashish Nagar
There are a few, the common ones. Just this weekend I was reading Hard Thing about Hard Things by Ben Horowitz. It’s very tactical. But there is a book which I and again, it sounds counterintuitive, but there’s a book on leadership in Turbulent times by historian Doris Kim Goodwin. She’s the historian behind the famous Nixon biography and so on. Her book on leadership was just talked to me a lot. It talks about eight American presidents and managing through World War II, managing through in Case of Lincoln, how bringing the country together and so on, and essentially managing through adversity. And in a startup, either you are a 3% company or 100% company or a 10,000 person company. There is adversity every day, every week, every month. Right. And if you don’t have adversity, I would almost say that you are not trying hard enough. Yeah. And so how these people, famous presidents and not so famous presidents, looked at life and created their life and managed adversity. 

Ashish Nagar
There’s a lot of learning for everybody there, especially startup honors. 

Brett
Yeah, nice. I’ll have to check that out. I’ve definitely read Hard Thing about Hard Things a few times and I think that’s one of those books that I just find myself going back to over and over again, but I haven’t read the other. Excited to check that out. 

Ashish Nagar
Sure. 

 

Brett
And now let’s switch gears and talk about level AI and dive in a bit deeper. So what’s the story behind the company and then what’s the high level pitch? What do you do for customers? 

Ashish Nagar
Yeah, so we start with the high level pitch. We are a contact center intelligence platform. If you think about every brand in the world. As you said, every company wants to talk about customer centricity, but you associate very few brands with actually being customer centric. Let’s say Zap or something old like a few years ago, and now Qi, Tesla to some extent, Apple to some extent, but we can have an argument, but if they have great customer service or not. But we can also argue that our cable service, you pick whichever one you have, and I pick whichever one I have, has terrible customer service, and in general, they’re not pulling their customers first. But every brand, every Fortune 500 talks about this. What level AI does is we are a contact center intelligence platform. We enables these major companies to actually review and analyze their customer service data at scale, bring insights out of that to feed to different parts of the business. 

Ashish Nagar
Product team, service teams, operations teams, sales teams, and also improve. Their customer service performance, improving their quality, improve automating some of the repeat workflows, and improving the life and quality of the agents who are actually working on these jobs. We don’t develop bots. We are not a chat bot or a voice bot company. We are a customer service intelligence platform. And the way the company came about when I left Alexa, we have a strong hypothesis that automating and enhancing the workflows of frontline workers, people unlike you and me, who have desk jobs, is a massive opportunity that is the largest part of our employment pool. Frontline workers are people like contact center workers, nurses, retail store technicians and so on. And we started building voice automation capabilities for some of them and then quickly realized that there’s a massive market on the customer service side where there has been not much new innovation in the last five to ten years. 

Ashish Nagar
And that’s what we are doing on the back of tremendous advancements in AI over the last half a decade. 

Brett
Nice. I love that. And are there any specific market segments where you’re seeing the most traction and adoption? 

Ashish Nagar
So we are focused on both B two B and B to C segments, but within B two C we see obviously a very big demand because, as I said, every brand in the world is struggling to differentiate, and customer service and customer experience is one of the biggest differentiators within that. Wherever there is complicated customer experience requirements, let’s say financial services, health care, insurance, travel to some extent, is where food delivery services in general is where we see the most requirement. Things which are super simple, like very simple retail customer service, for example, have fewer needs in these areas. But overall, the market for these technologies is tremendous. Such a big category within software that we are often surprised how big can this become? 

Brett
And you mentioned category there. How do you think about market categories? Are you trying to carve this out into a totally new category that you’re going to brand or is this chipping away into an existing category or yeah. What are your thoughts there? 

Ashish Nagar
We think that over a period of time there will be consolidation in some of these product features. And right now there are a few product features under which our work falls, but over a period of time it would become Customer Service Intelligence, Contact Center Intelligence. But currently we do chip away from a lot of existing categories in the customer service space. So real time agent assistant. Real time agent coaching is a category. Workforce optimization or quality management systems as a category. And then conversational analytics is another category which we take budget from, but we think medium term, these things should consolidate and they should consolidate into some supply customer Service Intelligence or Contact Center Intelligence category. 

Brett
And I saw on your site that there was a forester report there, a piece of forester content. Are you actively working with the analyst firms to try to shape that market category? 

Ashish Nagar
Exactly, we are, yeah. We’re closely working with Forrester and Gartner both to learn from them, obviously, and also, as you said, spread the word in the enterprise and broader community on the work we are doing and what is ahead for the space in general. 

Brett
That’s interesting. Yeah. I see a lot of founders that I talk to are split on that. Some of them say I hate Gartner, I hate Forrester, I don’t want to go through the analyst firms. And then others say, you can’t really go enterprise unless you do it. It’s just the way things are. So it sounds like you’re more in that second camp where you need Gartner, you need Forrester, and you need those relationships for enterprise. 

Ashish Nagar
Then we are, as I said, channeling my inner jeff. We are not in any camp. We are in the customers camp. Right. So the thing was when our customers told us like, hey, we check out Gartner about these things, they were like, sure, if you check out Gartner, then we are in Gartner. So that was a simple answer, to be honest. We were increasingly we had customers leaning in on forester and Gartner and that’s where we are partnering and we have a lot of respect for those folks and look forward to successfully partnering with them. 

Brett
Nice. That makes a lot of sense. And that’s, I think, helpful advice for any founder. That is. Yeah, don’t be on either camp for the customer. 

Ashish Nagar
Yeah. And in general, one of the other things is like, oh, are you an AI company or not an AI company? We are in no, we are AI and not AI or whatever it takes to win customer scam. Right. Because there’s a lot of noise about hey, but yeah, I’m sort of digressing here, but that generally is our thinking. Yeah. 

Brett
Nice. Makes sense. On that note of just noise, how are you standing out right now with AI? I feel like ten times a day about a new startup, how AI is going to change everything. And it seems like that story is just becoming almost commoditized, it seems. So how are you standing out? Are you really leaning on your background and the fact that you come from this space and you’ve done a lot of work here, or what are you doing? 

Ashish Nagar
Great question. So this is like now we are at the stage when we raise up series B and so on, and we are well funded. We have to stand out with customers we don’t have channel partners and people who we do business with. Right? So for there to stand out, it’s very simple. What problems can be solved for which no one has? And that is where the rubber hits the road on technology. And we are from a bunch of our unique innovations, ten X better than anything else out there. And so that is frankly simpler because of the quality and breadth of our product and our deep innovations. To your point about standing out in, let’s say, the investor community or in general, the broader messaging of it, there are a couple of things. I think they are again focusing on the specific problems we are able to solve, number one. 

Ashish Nagar
Number two, the founder story. I think at least investor, the founder market fit is really important. How that manifests is that because I saw a lot of what Amazon Alexa was doing at scale with AI. I have a unique sense on which problems to solve, where to invest in, where not to invest in, to be able to as a small company, to be able to differentiate us. The last thing I would say, honestly, we talk to hundreds of customers every month. The Twitter noise doesn’t really impact them. What impacts them is, again, like, can you solve my problem for which I have $100,000 budget any better than anybody else. Whether it uses generative AI or some other AI, I don’t care. Often customers don’t even know the difference between the two. And I’m not saying because they are ill informed, it’s just because it’s not their job. 

Ashish Nagar
They’re so busy solving their business problems, right? And it’s our job to create that magic with any technology possible and to solve it for them. Now, for example, to give you a sense, were using Generative AI for many years before it became cool. And there are many other techniques which we use which just solve customer problems. And yeah, maybe we should have created more noise about it, but we believe in putting the customer first. 

Brett
And because of your background as a technologist, I would guess you didn’t spend a lot of time in marketing. But I’m looking through your site. You guys have done a very good job in marketing, and it looks really good. So what have you had to change internally as a founder to go from a technologist to understanding marketing or at least hiring a team and building a team that gets marketing because like I said, it’s really great. 

Ashish Nagar
I appreciate that. It’s actually a pretty lean effort and what we are trying to do is create content which genuinely helps customers. Again simple things which can help our customers make better decision and then we have a lot of work to do on positioning ourselves as an AI force, SaaS product and so on. And we are working on that. But in terms of the change as a founder, I think one of the key elements is constantly reminding ourselves that we have to educate the space. Either if you are creating a category or you have a ten X better product, we have to educate the space and put out our message. The onus is on us and no great technology gets self discovered and self launched and self implemented. Other benefits like have massive social effects and so on. So our focus has been on targeted marketing to our customers and creating very useful content. 

Ashish Nagar
Those are two things which we are pursuing and it’s fun to do that because again, it comes very organically because you learn something, you have a great insight, you want to share it with the broader community as opposed to forcing yourself. Right? So that has been our approach. 

Brett
Nice, I love that. And going into your more forward looking of where we see things headed here, I know you have on your site the future of customer support at your fingertips. So let’s zoom out five years from now what’s the future of customer support look like both for the organization delivering the support and then also for the consumer. What’s the consumer experience going to be like? Hopefully better than today. 

Ashish Nagar
That’s a great point. So a few things which I’m pretty sure of, which is kind of to an earlier .5 years from now there will still be humans doing a lot of work on the side. If I had to pick a number, somewhere between 50% to 70% of the work would still be done by humans and there will be 30% to 40% automation based on the category and so on. When I say automation. So let me paint a picture for you, a customer version and being able to really talk to a reasonable voice bot or a chat bot, which with a human like conversation and not that not to the clunky systems. We have IBA systems, IVR system these days, press to get to a customer service rep, whatever right, have a real conversation. And then as soon as it becomes a business decision, as soon as it becomes a real technology troubleshooting or any other service requirement, there will be a human in the loop who will take the more expert calls. 

Ashish Nagar
The other piece where I think the demand, the feedback loop will become much faster because AI will be able to pick up insights on what’s happening in your where are customers not happy and so on much faster and much more accurately and act on it either on your behalf or help the customer teams help on it. And then also the accompanying products will be faster because if you have great fast feedback, you can improve your service and product much faster. Those three or four things I can clearly predict. Like one of the things, Brett, which people don’t appreciate, even with the rapid changes in AI which welcome, and we are really obviously optimistic about these changes, but people don’t appreciate how hard it is to have human dialogue. And I’ll give you a quick story to link to the vision. So the team I used to work for at Alexa, the team called Alexa Prize, and our goal is still an active team project to have talk to a human on any social topic for 20 minutes, okay? 

Ashish Nagar
On any social topic, as if you walk into a bar and you start a conversation with somebody. And Alexa right now can do that for less than two minutes, if that. And it’s really hard, and that links back conversation on automating customer service. It’s really hard to have human like conversations. There are so many unsolved science problems in there, problems which need to be worked on. 

Brett
That makes a lot of sense. And that was the next question I was going to have for you, is do you have thoughts or just concerns around this displacing workers? I think that it’s very difficult to talk about AI without the technology critics coming up and saying, okay, well, it’s going to put everyone out of work, so what are we going to do? What are your thoughts there and how do you approach that conversation? 

Ashish Nagar
I think there will definitely be some displacement, but we have the strong thought so level AI’s vision. And our motivation is to augment human productivity to make them ten x better. And we think the right balance is people like you and me and workers and humans doing the job, but in a five X more efficient way, ten X more efficient way, hundred X more efficient way with AI augmentation. And that’s the best synergy in our view, especially in the space we work in customer service. In some other spaces, like robotics or like just today there was some news on Amazon replacing warehouse workers with robots. There will be heterogeneous automation, but tasks where human ingenuity and human creativity is required. Now, some people can argue, like with the stuff which we are seeing on generative AI models in imaging, for example, right, with stable diffusion and so on, that the creative tasks are being displaced first. 

Ashish Nagar
I think that it’s limited. I think it’s like limited to some areas. There are a lot of areas where the creative tasks are here to stay for a long period of time, and their AI will be augmenting human productivity. There might be some pockets where there’s larger displacement. But I don’t see it as big a threat as most people make it in the media, to be honest. 

 

Brett
Well, that would be the first time the media has ever said anything that gets people scared. 

Ashish Nagar
Right, exactly. 

Brett
As a consumer. Let’s zoom out five years from now. Am I ever going to have to wait on hold for an hour again? That’s my example of I was calling United maybe two weeks ago and I think it was 45 minutes to an hour before I was finally able to get someone on the phone, which as a customer just seems so insane. Do you think that’s going to exist in five years or is that going to be something that we tell our kids of back when were young, we used to have to sit there for an hour waiting on customer service? 

Ashish Nagar
No, I’m with you, man. I think that should go away. And if it doesn’t, and I think it’s going away, leading brands even now, that should go away. There’s no reason for that to exist either with human productivity or without it. There are so much faster, better, elegant ways of handling that traffic and servicing customers than that. I think that should definitely go away. And especially for first order request, there are so many interesting things which you can do in terms of voice automation, text automation, and then creating digital self service loops and things like that stuff should go away. And I’m surprised that United hasn’t even done that yet. One of the other underlying things which we see, and I’m sure your other guests have talked about this, the underlying It infrastructure right. In a lot of these places. So to give you a sense in the contact center space, you might be surprised hearing this. 

Ashish Nagar
In the customer service or contact center space. Even now, more than 75% of all contact centers are run on premise systems, which is they are not on cloud, which means that when a phone call goes in, there is a physical PBX system which is routing those phone calls as opposed to a cloud telephony system which is routing phone calls in a cloud environment. Right. What does that mean for your question? It means that in a cloud environment you can have a very simple interconnectivity with AI with very different things and newest technology, but with brands and companies who don’t have that, they’re struggling to modernize their customer service stack. Right. And that’s where you can see some of the laggard experiences here. 

Brett
And why do you think that is, that they’re so slow to adopt? It feels like every industry has moved to the cloud, or at least that’s what it feels like from what you read. Why is this space so slow? 

Ashish Nagar
That’s a good point because things are changing rapidly and you’ll see that in the growth of companies like Five Nine, Twilio and Genesis and so on. But I think it’s a combination of that inertia and just the weight of changing and uprooting these highly complex systems when complexity is keeping your lights on. So it’s not as if you can’t take calls anymore. It’s keeping your lights on, but it just doesn’t make you ready for the modern customer. Right. So in boardrooms and CIO rooms and customer service executive conversations, that changes rapidly happening. But clearly such a large market, it’s still not fast enough. You’ll be amazed and I’m sure you see that the amount of work it takes to have a digital transformation effort like that, it’s ridiculously big and it takes five years or three years sometimes to replace some of this stuff. 

Brett
Yeah, makes sense. I think from the outside it always seems so easy, right. You just digitally transform and you’re good to go. But the reality is it’s a big undertaking and I’m guessing takes a lot longer and is more expensive than anyone anticipates. 

Ashish Nagar
Yeah, and no one in the ecosystem is incentivized to make it faster and cheaper. Everyone wants their slice of the pie and add their value and kind of it’s, so it goes. 

Brett
Makes a lot of sense. And on the conversational AI point here, what excites you most? Are there any specific developments? Not talking about what you’re doing as a company, but just the technology in general, what excites you the most about what’s ahead? 

Ashish Nagar
Man, there are so many things. So I do think like this whole thing about generative AI about which in case of conversational systems, it’s machine’s ability to create language, right? Machine’s ability to write stuff and create pros or create content. That’s very exciting. And that’s one area. There are a lot of unsolved problems here. How do you like in our conversation, for example, Brett, it’s very simple. We talked for about 25, 30 minutes. No AI system can absorb this conversation and can actually have complete memory of what happened here from a perspective of modeling this conversation or be an active participant in this conversation, given whatever else has happened in the past. Right. So being able to model conversations and mimic human dialogue over a long range, not in short, one, two instances, but over a long range is a very interesting problem which really excites me. 

Ashish Nagar
And we’ll see what happens there. And the other thing which is an unsolved problem is reasoning itself. And what I mean by that is these language models and all the stuff which is coming out, they don’t have their own reasoning. They look at data, like billions of data points in some cases, but they are not like humans do, learning about cross disciplinary things or having diligence about cross disciplinary things. That is a very interesting area. So I would say those three things, the generative stuff, modeling conversation more broadly and bringing intelligence, like cross disciplinary intelligence and reasoning more broadly, that’ll be really cool. And I’m sorry if I’m going towards. The geeky end of this. Sorry. 

Brett
That’s our audience. They like the geeky stuff, so they appreciate that, and I do, too. Last question here for you. So I know we talked about the future of customer support, but if we talk about the future of level AI, if we zoom out five years from now, what’s that big picture vision for the company? 

Ashish Nagar
That’s a great one. So the answer is sort of in. One of the questions you asked about our vision is we want to augment human productivity, starting with the contact center for tens of millions of Americans to begin with, and what does that look like? So using AI to automate, analyze, and improve enterprise workflows, starting with the customer service function, because customer service is the largest store of unstructured data within the enterprise. And then look at other functions where we think there’s a massive opportunity for these capabilities. And if we look back in five years, we would have influenced, hopefully, a few million lives and made them happier, more productive, and more fulfilling with our technology. And what’s not to like with that? Because just in customer service alone, Brett, there are 5 million Americans who work in customer service. 5 million. Why not, right? 

Ashish Nagar
So if you can improve their quality of life by 20%, 30%, 50%, that’s very motivating for all of us working on this mission, and that’s what we look forward to do. And as a result, improve these brands and improve these businesses to sell more to service their customer better, creating overall happiness, that’s what motivates us. 

Brett
And eliminating the 45 minutes hold times with United, please. 

Ashish Nagar
That’S at the top. Someone for United is listening. They know who to reach out to. 

Brett
Yeah, exactly. All right, man. Well, unfortunately, we’re up on time, so before we wrap here, if people want to follow along with your journey as you build this company out and make this vision into a reality, where’s the best place for them to go? 

Ashish Nagar
Yeah, we are on Twitter at the Level AI and LinkedIn at Level AI. Or just go to our website, the Level AI. Awesome. 

Brett
Well, thanks so much for your time. This has really been one of my favorite conversations so far, so really appreciate it and look forward to keeping in touch and wish you the best of luck. 

Ashish Nagar
Thanks, Bert. 

Brett
Take care.