Hello, and welcome to the Nvidia AI Podcast.
I'm your host, Noah Kravitz. Customer service has evolved from a counter or back office where angry shoppers go to complain to an integral part of modern business.
From in-store and over-the-phone help to cross-channel online support, customers expect support and businesses are pouring time and money into new ways to up their service games.
Satish Mandalika is founder and CEO of Drisham AI, and he's speaking on a panel discussion here at GTC 2019. exploring how AI is reshaping customer service in key industries and how AI is making it more customer-friendly.
Satish, thanks for taking the time to join the AI podcast and welcome.
Thanks for having me, Noah. So tell us about Drisham AI.
What do you do? So Drisham AI is a Bay Area based startup. that is applying artificial intelligence, specifically computer vision, to simplify customer support for manufacturers.
So the idea is as follows. The customers of these manufacturers that we're building this platform for... could just take pictures of the product that they need support for.
They don't have to get on a phone call and spend 30 minutes trying to explain a product.
Our platform will take care of that automatically.
All the customer would have to do is use their mobile device, their tablet, etc. to take a picture of the product that they need support for.
We're sitting in a conference room here at the show recording and there's a bunch of gear on the table.
So let's pick, there's some headphones. So I would have an app.
And I would just snap a photo of the headphones and that would connect me.
And you said manufacturers. So this isn't to the retail outlet.
I'm going straight to, I don't want to call out the name of the headphone maker because these headphones were great, but it would take me straight to the manufacturer, to their support line.
That's exactly right. And there's a reason why we picked the manufacturers.
So if you think about the home improvement domain, most of the products that get bought you know, either from Home Depot, Lowe's or Amazon, the lifetime warranty for that product is held by the manufacturer.
Okay. So eventually, if the product breaks down, you would eventually get on a customer support call with the manufacturer to see if, you know, that they are offering replacement parts for free or what have you.
So from that perspective, what ends up happening is the customer support calls for these manufacturers are through the roof.
And that leads to long wait times and unhappy customers.
Yes, I think we can all relate. That's right.
And so that's where we're trying to serve the manufacturers by making it easy for end customers like you and me to be able to get to the resolution of the problem for the product that we need support for faster.
Right. No, that's great. And so you mentioned the home improvement domain.
Is that one of your key industries? That's right.
So we've built a platform that could be applied across multiple domains.
But home improvement has been our beachhead, our go-to-market, because that's where we saw an immediate problem that could use our platform to solve the problem elegantly.
And how much I'm thinking when, you know, I picked headphones at random.
And then when you said home improvement, I immediately thought to. oh yeah, my dryer last year.
And then a couple of years ago, our oven, you know, we replaced our oven and How many of the use cases are direct manufacturer-to-consumer work?
Or, you know, in this industry, are you going through kind of contractors and service professionals as well?
Or how does that work? Again, like I said, our focus has been manufacturers to begin with, and we made the conscious decision to go after manufacturers because we're trying to solve this problem one manufacturer at a time.
Now, that helps us to get to the critical mass of data where we build that intelligence across product lines in a domain.
And then we can turn around and help the consumer directly.
So that's our ultimate vision. But we're currently focusing on manufacturers.
Right. And so I would assume computer vision comes into play in identifying the product that I've taken a picture of.
You talk a little bit either about that or about other ways you're using computer vision, machine learning, and AI in your stack?
So it's been a great time for startups involved in computer vision.
Computer vision technology has matured in this particular wave of AI to a point where it can solve really interesting problems.
And we're starting to see results of that already.
So self-driving car is the best example out there.
That's been possible because of the level of accuracy that computer vision based on deep learning has gotten into.
Other applications are identifying rare form of cancers, all kinds of interesting problems that we're solving using And so we decided to take this technology and apply it to customer support because of the problem that we saw.
And that got us started on Drishyam. And so where do you see the customer service industry, if you will, headed?
And what AI's role, you know, kind of is or will be evolving into going forward.
I think customer support is ripe for disruption.
I don't think... We have to be on a customer support call for an hour or more to get support for products that we're having issues with.
And it doesn't have to be at a certain time of day.
AI will let us get into support in a much more elegant way across multiple channels. a time of day or night that's convenient to the customers.
Right. I know I, in particular, and this is retail, not manufacturer, but if I can find... the way to do it, not on a phone call, you know, and even if it's, and to your point, I could do it late at night.
It could be something where I get into the queue and, And when I'm waiting for a bot or a person to come online to help me, I can do something else, that kind of thing.
So, you know, clearly from the customer end, it makes a ton of sense.
Do you know, when I go on and I mentioned, you know, I don't know if it's a bot or a person, any idea how often when we think we're chatting with a customer support rep, it's actually an an AI-driven bot helping us out?
Yeah, I'm not exactly sure about the numbers, but if I were to take a guess, I'd say... six or seven out of 10 times, there's a chance that you might run into a bot.
Because I think... Again, like I said, the industry is headed to a place where the solutions being offered in the form of customer support are now a combination of bots and humans.
And so they can take the rote work out of helping you identify products, all that stuff, and make the customer support person more effective in his job.
Productive. Right. Right. So let's switch gears a little bit.
We're talking to Satish Mandalika. He's the founder and CEO of Drisham AI.com.
They're using AI to, I'm going to use your words, disrupt the customer service industry, working directly with the manufacturers.
It's very cool. It makes a lot of sense.
As a consumer, you can snap a photo of the item you're having a problem with. send it off to the manufacturer and get support directly that way.
But you've got an interesting background.
And we were talking a little bit offline, kind of a...
Bit of a kindred spirit in terms of a nonlinear career path, if you will.
You've got an entrepreneurial bent. You've worked in healthcare, e-commerce.
You've been in-house at places. You've started your own companies.
So maybe kind of walk us through some of your journey, how you wound up Starting Drisham AI.
And also, if you could, how you've seen machine learning and AI kind of develop and evolve and play a role in your own business? career path?
I think I made a conscious decision to get into entrepreneurship.
In my mind, there isn't a better place on this planet to start a company than Bay Area.
Because I think the Valley has gotten it down to a science when it comes to building and scaling startups.
And so. And I should mention, I don't know if we did, that Drisham is a Bay Area company.
That's correct. Thank you so much for having me.
I've built a career in writing and building enterprise software for large companies, healthcare companies, financial companies, etc.
And so 10, 15 years back, I got my MBA from high school of business.
I signed up for a class in entrepreneurship.
And absolutely fell in love with that concept of creating jobs and money out of thin air.
Right. And so right out of that program, I decided to at some point start my own company.
And so I come from a very strong enterprise background, having done digital strategy for large banks and healthcare companies.
And so I wanted to apply that. to solve one of my pet peeves, which was I wasn't handy with home improvement tools.
And so when my home came up for maintenance, I used to struggle with it.
You know, these constant runs to the Lowe's and the Home Depot is doing competitors and shopping with Amazon.
I felt like there was a better way of shopping in this domain.
Right. And so that got me started in my chance meeting with my partners in crime, Aditya and Sandy, got us thinking about solving this problem more elegantly.
Not just for myself, but maybe for... an entire market out there.
And so that's how the seed of Drishyam was sown.
And how long ago was that? We started this company about two years back.
Okay. And how's it going? It's been going great.
We've been associated with NVIDIA. We're part of the NVIDIA Inception portfolio.
And it's been great. There's, like I said, a huge market for AI, a lot of startups doing interesting stuff with AI.
When we went to market the platform, the manufacturers that we were targeting really liked the concept of what we're doing.
And we were easily able to sign up to the largest faucet manufacturing companies in the world as our pilots.
And so we're getting ready to convert them to our paid clients.
And we've got a healthy pipeline of customers that we're talking to.
So we're excited about what the future holds for Drishyuk.
Good. Well, as somebody who can also share in the willingness and frustration of trying to do repairs around my own house, sometimes to my wife and children's chagrin.
Yeah. I, you know, I'm excited for you guys to keep growing and making things easier for the rest of us.
That's fantastic. We talked a little bit about your entrepreneurial background.
You mentioned having kind of been through these three boom and bust or Two boom and bust and one now boom cycle, right, in the Bay Area tech scene.
And we get a pretty wide... you know, variety of people listening to the show, but certainly folks who are, whether they're younger or not, they're interested in, either making a career change, getting into AI, certainly starting their own ventures.
Drawing on your wisdom, being through these cycles and being very hands on with everything, what, if any, advice would you give to folks, whether about entrepreneurship or about writing these cycles of boom and bust, and then certainly the current moment with AI?
So as it pertains to AI, I would say folks that are trying to get into AI or trying to have a career in AI...
I would say embrace it with open arms. AI is here to stay.
And I know I've heard a lot of doomsday prophecy that it's going to end the world or robots are going to take on the world.
I think AI is here to better our lives, to improve all different aspects of lives. embrace it with open arms.
I think more jobs will be created, then more jobs will be destroyed.
I know people are worried about AI destroying jobs, And so I think there's a lot of optimization that's going to happen where...
AI is going to better our lives before we think about job elimination.
And I'm pretty sure there'll be more jobs created than jobs destroyed. there are different aspects of it.
Obviously there's the technical aspect. So it's a great time to get into coding and all that.
And then, uh, I know there's a lot of awareness now, much more than there was 20 years back. folks that are trying to get into technology in general, I would say give it a shot.
Things have dramatically changed in the last 20 years.
Did you have a Comp Sci background before getting into AI?
I did. I did. Did you pick up AI skills on your own?
Did you go back to university? How did you develop those chops?
I did. So again, I do have a degree in computer science.
That said, I haven't I'm kind of rusty with my coding skills.
I haven't coded in a long time, especially since I switched to product management.
But that said, I was always curious about technology in general.
So it actually started with me signing up for a course in General Assembly.
So a quick shout out to General Assembly.
I think they did a fantastic job introducing AI to beginners.
And I think the entire cohort that I took that class with absolutely loved it.
And that actually got me into machine learning and then eventually into deep learning.
There are a lot of avenues out there and now it's even getting better.
I took a classroom course, but there are a lot of online courses that you can sign up for. at your own leisure and try to explore these technologies.
No, it's always interesting in the course of doing the show and interviewing people.
It's interesting to talk to folks who...
Have those background, you know, a background in something related and then picked up an online course or, you know, realized they had an old gaming PC, not that old. but a gaming PC that like, oh, wait, that GPU, you know, I can do some algorithm stuff with it.
And so it's a fascinating time. So looking back on your time, not just with Drisham, but again, tap into your knowledge and experience. unique purview on the Bay Tech area over the past couple of decades.
Is there anything that's happened along the way that when you think back was kind of surprising to you? whether technological development or something in the marketplace or entrepreneurship or specific to AI.
I never expected I had to go mainstream this quickly.
I've been... surprised and blown away by some of the results that we're starting to see across the industry and what I've been seeing in our product.
To be honest, when I started in Drishyam, I was kind of skeptical about how it's gonna pan out, But the results have absolutely blown my mind.
So that's why I keep saying AI is here to stay and make best friends with AI.
It's not going anywhere. Sage advice. What's next looking forward for Drisham?
And also you mentioned, it's something we talk about a lot, this forget the doomsday prophecies, but AI bots, if you will, and humans can really work together, you know, and that... notion of kind of mastering like, well, this is what the computer can automate and leveraging its strengths.
And then this is where the human creativity and management and all that stuff comes in to kind of guide it.
Where do you see this all headed in the next few years, 10 years, whatever timeframe kind of makes sense to you?
Sure. So I'll start with Drishyam. So it's been a busy 2019 for Drishyam so far, and I suspect it's only going to get better or worse.
Stick with better. Yeah. Well, and there's a lot of work coming down the pipeline.
We're, like I said, in the process of converting all these pilots into paid clients.
There seems to be a lot of demand and that's going to. keep us busy for the next few years.
We're also raising a seed round. So if there are any angel investors or investors... listening in.
Love to talk to you guys. So our vision. is to eventually develop the platform to a point where, like I said, we should be able to completely eliminate customer service.
You should be able to get support for any product that you need support for just by pointing your mobile device at it.
And platforms like ours will then help you identify the products, troubleshoot and even order parts and all that.
So a very streamlined experience at your convenience, you know, whether you're on a Facebook Messenger, a mobile app, chatbot, whatever it is, it's going to be pervasive and the technology is there to help you out.
Now, will you also automatically order a skilled professional and medical support when I'm trying to... fix my own HVAC system in my house and something goes horribly awry, that's my fault, not yours.
So it's funny you mentioned that there are startups already starting to do that.
So they're taking it to the next level where to your point, identifying skill people, who can actually help you come fix the problem that you're having.
Well, I, you know, joke from experience, luckily not the medical professional part, knock wood.
But, you know, I have talked to some folks who do field service ops and that kind of thing.
And, you know, using... you know, conceptually similar technology.
And it just made me think like, you know, going forward, there's just to be able to integrate everything. for the system to kind of help, you know, when you need whatever it is, a part, someone with more skill, you know, whatever it is.
Absolutely. It's fantastic. Satish Mandalika, the company is Drisham AI.
If people want to find out more about Drisham, The work you're doing, companies you're working with, or even your own background, there are places they can go online.
So you can visit us at our website, www.trishyam.ai.
And we have a Twitter handle that's pretty active.
It's Drisham underscore AI. Okay, so for all the angels out there listening.
Hit you up, slide into your DMs, talk about that seed round.
Thanks for the plug, Noah. Appreciate it.
Well, we'll see if that gets passed corporate.
Um, Satish, thank you so much. I, again, you know, it's fascinating times and as a person who, um, brute forces my way into finding my own little hack so I don't have to sit on customer support line waiting.
I applaud you and I have my fingers crossed for much success for what you're doing.
Thanks a lot, Noah. Thanks for having me.
My pleasure. Enjoy your show. Thanks. Thank you.
¶¶
