Auto Collabs
Collaboration through connection.
Hosted by Paul J Daly, Kyle Mountsier, and Michael Cirillo, Auto Collabs is your connect point to the human side of the retail automotive industry: what motivates its leaders, how they navigate change, and what keeps them pushing forward. It’s beyond-the-business-card conversations with real people powering dealerships, technology platforms, and everything in between.
From the team behind the More Than Cars movement, this podcast is built on one big belief: thriving people create thriving businesses. With candid conversations, industry insight, and just enough unfiltered banter to keep things interesting, Auto Collabs delivers authentic stories that inspire real collaboration.
// Auto Collabs is produced by Automotive State of the Union (ASOTU). Learn more at https://www.asotu.com
Auto Collabs
Training AI To Fix The Little Things with Sanjay Varnwal
This episode of Auto Collabs gets seriously smart as Paul, Kyle, and Michael welcome Sanjay Varnwal, founder of Spyne. With a background that includes time at Amazon and deep roots in AI-driven product development, Sanjay shares his journey from general e-commerce merchandising to building ultra-specialized solutions for the auto industry. His company’s obsession with perfection—like making sure side mirrors don’t vanish in digital photos—isn’t just nerdy, it’s the kind of tech precision that can radically improve dealer efficiency and customer experience.
Sanjay also breaks down the evolution of agentic AI—technology that doesn't just talk, but acts on behalf of dealers to optimize pricing, image performance, and even ad budgets. It’s the next level of automation, where your digital teammate gets smarter (and more autonomous) over time. Whether you're deep in the tech weeds or just trying to understand how AI can lighten the load in your dealership, this convo offers sharp insights and a clear look at what’s next.
Takeaways:
0:00 – 🎙️ Intro banter: Meta Ray-Bans, Weezer dreams & Wiggles hype
1:48 – 💡 Introducing Sanjay Varnwal from Spyne
2:28 – 🤖 Sanjay’s first AI experience at Amazon Go
3:39 – 🧠 How AI’s been around long before 2022
5:02 – 🎥 The evolution of AI in media and merchandising
7:40 – 🔄 How AI is changing human behavior and productivity
9:45 – 🚗 Why Spyne focused on auto merchandising
11:30 – 🛠️ From general e-commerce to auto-specific solutions
13:05 – 📸 Solving hard problems: lighting, stabilization & studio-quality images
15:46 – 🪞 The vanishing mirror problem and how Spyne fixed it
17:40 – 🧑💻 Building an AI-powered teammate for dealers
19:04 – 🧭 What is agentic AI and why it matters
20:25 – 📊 Real-time merchandising optimization and ad performance
21:01 – ⚙️ Letting AI handle repeatable, data-driven tasks
21:40 – 🙌 Final thoughts and future of AI in automotive
Connect with Sanjay Varnwal at https://www.linkedin.com/in/sanjaykv/
Learn more about Spyne at https://www.spyne.ai
⭐️ Love the podcast? Please leave us a review here — even one sentence helps! Consider including your LinkedIn or Instagram handle so we can thank you personally!
We have a daily email!
✉️ Sign up for our free and fun-to-read daily email for a quick shot of relevant news in automotive retail, media, and pop culture.
🎧 Like and follow our other podcasts:
Michael, you have some new glasses. They talk to me.
Unknown:This is Auto Collabs.
Paul J Daly:I mean, I thought, I thought, Well,
Kyle Mountsier:Michael, let me break it to you. Those are just the voices in your head. They've always been there. That's just what
Paul J Daly:they did to make you feel better about it. It's a new therapy. They give you glasses. They say, talk to you so that you think it's, oh, I'm normal, but you're not. I actually thought you were trying out to be the new lead singer, Weezer.
Michael Cirillo:That would be, man, that you just brought me right back to the seventh grade.
Paul J Daly:I mean, I was pretty much trade and I would, yeah, I mean, you know, guys, if Weezer calls me one day you're gone is like, say, hey, we need a new front man, or we need a drummer. Like, you're just never gonna see me again. I'm just
Michael Cirillo:pretty much any band. Like, there's, I have, really any band. You
Paul J Daly:trade us for, any band.
Kyle Mountsier:Well, he's like some death metal van band. Done
Michael Cirillo:these shorts, Jonas Brothers. Now the audience is throwing me under the bus, bro, I don't think I'm not missing this. You got the better bids? They'd rather be your
Paul J Daly:friend. Listen, listen, it's
Michael Cirillo:just exposure. When I'd take any band, at this point, I would take like, I'd join the wiggles.
Paul J Daly:Go hard with the kid. Strangely enough, that is probably the most likely band to call you, because you would make an amazing wiggle,
Michael Cirillo:just amazing. I'd be the hype guy, like the mighty, mighty boss tones. I wear a gray suit in the
Paul J Daly:back with with Jack Black vibes, and you're dancing,
Unknown:just kicking everywhere.
Kyle Mountsier:Well, all that has nothing to do with the fact that we are hanging out with Sanjay Vaughn, wall from from spine, AI, talking AI. Man, we're talking AI. We're gonna be way more intelligent, hopefully, than us talking about the wiggles. But if you want to listen to the wiggles,
Paul J Daly:no, he's wherever you want to, Sanjay is easily going to be bring the collective IQ average up by many points once he enters this room.
Kyle Mountsier:Yes. Why? Look, we really hope you enjoy this conversation with Sanjay.
Paul J Daly:Sanjay, how are you today? Thank you so much for joining us.
Sanjay Varnwal:Hey, thanks for inviting me. For the for the for your show. Paul Kyle and Michael,
Paul J Daly:yeah. So, so I'm sure to talk a lot about AI today, but I, I think this is going to be the question I start asking people, whenever we're talking about an AI topic, tell me about the first experience you ever had with an AI product.
Sanjay Varnwal:Oh, that's tough. I mean, no one has asked me this question before.
Unknown:Yeah,
Sanjay Varnwal:nailed it. So I'm remembering few of the things I used to work at Amazon in turn, 14 or 15 or so. And there, I used to travel to Seattle often, and I saw Amazon doing a pilot of a store called Amazon Go at that point of time. And it was one of its kind. When it was launched. You can just check and go inside the store. You don't need to do anything. Just pick up goods from there and go out and everything is like, checked out for you, payment is debited, everything. So Amazon was doing that particular pilot. Of course, the smaller AI use cases I would have seen in the past, but that was the first big, working, practical use case that I saw on the ground,
Kyle Mountsier:which, yeah, I think that that's interesting, because for the majority of the world, AI didn't really exist until, I don't know 2022 in their brains, right? But these big corporations have been dealing with AI since, like, really, the early 2000s there were already inklings of like, how do we create things that can do stuff on its own, right?
Sanjay Varnwal:Absolutely, although Google recommendations and all these things like, there is an element of AI behind it already, right? So, go
Paul J Daly:ahead. Go ahead, continue. Yeah. I mean
Sanjay Varnwal:the AI in terms of data analysis, data analytics, some of these tech has been existing for long, long time, but computer vision, where you see, like the magic happening in front of you is something that that started getting into prominence in the last like seven, eight years or so, yeah, just blown it up.
Paul J Daly:When was that movie that came out with that, that little kid, I just think of him as like the sad little boy, because he's got those sad eyes. Haley, Joel Osmond is and the movie was called AI, Oh, yeah. And I feel like that was the first time that I even, like, thought about artificial intelligence, yeah, right. And that was probably, I mean, I gotta say that that movie had to come out 15 to 20 years ago. But if that movie came out, I bet, I bet there was something going on in the tech world that somebody was like, You know what? We see this coming. We're gonna make. A movie about it, and now it's like movies reality.
Michael Cirillo:I love guys like Neil, degrasse, Tyson and stuff, though, because they bring everything back to its most rudimentary form. They're like, you know him? He's like, Oh, you want flying cars. We have those. They're called helicopters. And Sanjay, you made me think of this because, you know, with Google, with what Google's worked on over the last, I don't know, 20 years, anybody that's, if you want light reading, go learn about Google's Rank Brain, because that's been around for quite some time, and that is, I mean, that is what Skynet might be made of.
Sanjay Varnwal:Rank Brain, fundamentally, is the core machine learning base on their search algorithms are built, right? So I haven't studied in depth, but I've heard that this is, this is the base based technology on which the entire Google search functions globally. So got it,
Michael Cirillo:and it's like super it's like the treasury of the United States of America, like it is ironclad. There are, there are like, Rank Brain, Google, Rank Brain offices all over the world that you know. For example, we heard through the grapevine that there's an office in a lab at the University of Alberta in Edmonton, but there's no Google signage, any like they don't want people to know what's going on.
Paul J Daly:I see, yeah,
Michael Cirillo:it's deep, you know? It's funny, Sanjay, like we were just talking. I saw this notification in Slack that slacks just released. Its AI to our to our organization, and so I started playing around with it. I'm curious. You know what your take is on, on, on this if AI, let me back up see my brain's moving so fast, maybe this is I need the AI.
Paul J Daly:It's those meta Ray Bans. You the questions it's asking.
Michael Cirillo:Yeah, John Stina is in my ear telling me what to ask. I saw a meme recently, and it said in the year, what was it in the year 2050, that this is how the generation will communicate? And he was basically like a caveman again, prompting GPT. He was like, Jeremy.
Unknown:Good thing to say to girl, you know, like, there's like that just morphed into GPT to the degree that we can't even communicate as humans anymore. What are you seeing? I mean, I know you're heavy into this space and software development and the implications and applications of AI. Where do you see? What's happening out there? What's going on?
Sanjay Varnwal:I think you said it right. I'm not too sure if that would be the situation, but potentially could be. But yeah, as you see things around the way, the behaviors are changing for all of us, right? So I mean, earlier, we used to think, and then you used to write or used to say anything, but now I see myself, my behavior is changing to be whenever I need to respond to anyone, I I ask chat. GPT, hey, this is the profile. This is me. This is the context. Can you come up with this direct, I mean very I mean the kind of response that that elicits some response from them, right? So, and it comes out with very well researched answer or the responses. So they're like, just one of the use cases. I keep using chat, GPT for a variety of use cases, and have stopped applying my brains to a lot of work. I think that, you
Kyle Mountsier:know, I think that there's different levels of usage. One is like, here's a here's an email. Give me a response, but you actually are doing some critical thinking in that thing that you just said, which is, like, here's, here's something that someone said, here's a little bit more about them, here's a little bit more about me. What I want is this, this, this and this, give me some feedback in order to do that. And I think that that's, that's the layer that I think that we have to be thinking about any of these tools with, is, how do I take my like, knowledge of the world, my knowledge of the context of the situation, inject into what I'm what I'm using, and put it in I'd like to kind of pivot the conversation into, you have a unique knowledge. You worked at Amazon, have worked at some of these larger companies, and you saw it fit to move into automotive as a vertical that you wanted to serve. Why was that the layer that, like you, you brought to the conversation that was, hey, I'm seeing all these things move in the world, and I'm looking at retail auto and saying that's where I need to see a fit,
Sanjay Varnwal:yes, yes. So we picked up merchandising as a core problem in the industry, right? So, amen
Paul J Daly:say that twice.
Sanjay Varnwal:So mostly like digital merchandising, it was the problem that I picked up in. The in this particular industry, working at Amazon and so many other e commerce firms, realize that sellers, day in day out, struggle with producing high quality merchandising. I mean, the bigger sellers can hire expensive photographers agencies, but the smaller ones, long tail, do not do that. And the photo qualities and the in general media quality is shit very bad, it's and that also was not helping them sell more products online. So the idea was, can we create something which is more AI first in its approach, understands what product has to be shot, how to shoot it like a photographer, and then process those images into something that Amazon understands, that maybe Auto Trader understand that. So these, like larger platforms, they have given the guidelines and the course shoots are attuned for those guidelines. Picked up. This worked in this idea, launched something, ran it for two years, and then figured out that if you, if you are going to do everything, will be average at everything, right? So real estate has to be shot differently, versus fashion, model versus the product, versus the car versus other other objects, right? And that's where we realized that, okay, we need to pick one category. And car was always a passion. I knew a lot of dealers in my locality, so assured them the product. They were like, very happy. We got it rolled out to some of the biggest companies in the country, in India, and then we saw that this use case was working like phenomenally well, and US Europe being extremely huge market. We thought, Okay, this is, this is the direction that we led to. So
Kyle Mountsier:you actually went at, like, just general merchandising. Could have been any type of product. First, is that what you're saying?
Sanjay Varnwal:Yeah, that we focused on energy to build, like, the best product in automotive? In
Kyle Mountsier:auto? Yeah, I think that that's, that's an interesting start and pivot. Because when you do look at Amazon and you look at these large third party aggregators that are that are very similar to auto, I I always, I always make this argument, like, everyone's like, I don't want to be on the third parties. And it's like, well, the rest of the world operates that way. So why wouldn't we too? I mean, anybody like Amazon, right? That's a third party aggregator of the products, right? But when you think about, you know, the global market, the global merchandising efforts, you know, Nike spends millions and millions and millions of dollars a year to merchandise their shopping results pages, right? Because they know that that next click is critical, right? And, and we kind of like, you know, wait 11 days till we get a car photod and online, you know, because it has to get through the shop or whatever, and then, and then it's maybe got photos. And we've wasted 11 days. And I think that that merchandising flow in auto, you know, obviously we have unique VINs. It's a little bit different than, like, one shoe that you can shoot one time and have available for six months. So it's a different type of problem. But I love that you're trying to solve it. What have been some of the hardest things that you've encountered along the way, when you're like, well, we set out to build it this way, but that was actually harder, or that that was a mistake or a thing that you didn't expect along the way.
Sanjay Varnwal:Yeah. I mean, the goal that we with, which we started was we saw Carmax merchandising in very depth, car max or Carvana merchandising in very depth, right? And understood that some of the largest companies are setting up those 100,000 worth of studios in every single location to do their photo shoot. And the idea was that, hey, small dealer would not be setting up that 100,000 worth of booth because of real estate constraint and maintenance constraint and initial investment, right? But they all want that kind of a catalog merchandising experience. Idea was, can we get every single dealer to that particular experience without having to invest in that studio? That was the idea that, I mean, it should be indifferentiable from human eye, whether the car was actually shot in the studio or shot outside. That was the benchmark that we were tracing and we launched the product. We, I mean, you can check some of these studios. These are very realistic, glassy kind of reflections of the car. When, when we produce outcomes, right? And so we, we create behind the scenes some 7080, computer vision models, focusing on reflections, focus focusing on stabilization, focusing on interior view generation, focusing on like, lot of these problems we understand, like, hundreds of parts of the car, and every part are tuned in a different fashion when we process the image for this car. So the more hard problem that we got into was, one was stabilization, because when you shoot a car and move around it, your hands are unstable like and you when you try to place it in the studio, your car will look wobbly because the studio is perfect. How do you make that car spin perfect in that particular studio? So that was one of the technologies that we worked almost like over a year to perfect with. Now, if you see our 360 spins, no matter how you shoot it. Outside, it will always be perfectly placed in that particular studio, very, very smooth. This is one problem. The second was lighting problem, right? So exposure, lighting, some of these, oh my gosh, yes. If you shoot in the open yard, like harsh sunlight or during the later half of the day, so you'll see all like one side is lit, other other side is dark. Some of these problems we kept on encountering. And so we we created those exposure correction, Color Correction algorithms that even the light on the car, and based on the studio's outlook, it will, it will, it will project that kind of a tent on the on that particular car, so it will look like that it was shot in the studio. Some of these things we did which, which were, like, really hard problems. Like, we stayed on these problems for months, six months to 12.
Kyle Mountsier:How long did it take you to make sure that mirrors didn't disappear? Because that was my favorite thing in, like, early even, like two, three years ago, like, it's just missing a whole mirror, right? Or, like, the tires are gone. It's like, who lifted this car and made it a hover car? Back to the hover car thing, right? Yeah. Like, why? Why was that an issue, and how did you solve it?
Sanjay Varnwal:Yeah, so. So, whenever you create any kind of computer vision model or any kind of AI model, it has to understand the shape and size of that object on which it is operating, right? So that is also one of the reasons why we went into automotive, because the shapes were defined earlier we were doing like multiple categories. So I mean, a human versus a bag versus anything, right? So everything has a different shape. So when we picked up cards, we trained our algorithms with, I mean, few millions of images which were custom trained by our own annotators, and we specifically focused on this problem like mirror was one of the biggest, hardest problems that we solved, mirrors and antennas. So yeah, the antennas for sure, the antennas, right? So these two were one of the toughest problems that we solved in background removal, as well as those open, open door shots that you take right so when, when things can go wrong. So some of these, like specific hardware use cases, we generated our own custom data in like hundreds of 1000s, and then trained our models to make that to work. Now. Now it works with like 98 99% plus accuracy. And you give it any kind of image, it will
Paul J Daly:work. I mean, you all do? You all do so much more than photos. You know, I heard that you're building like an AI powered teammate, and I'm assuming this kind of, like, delves into the whole realm of agentic AI and so. So tell us what you're excited about. Like, what does the roadmap look for look like for what you're building right now?
Sanjay Varnwal:So, so we are. We are. I mean visualizing us as a company which will build AI first experiences for automotive industry, right? So merchandising was one of the use cases that we targeted. And merchandising natural extension is to go in other use cases of inventory, which is, how do you price, how do you inspect? How do you syndicate the car? Well, so this is like one, one spectrum on which will go very deep and build AI native experiences. The second, second part where we are now getting into is the this entire customer interaction of the dealers, right, using agent AI. So whatever we are doing now, we are looking to bring in agent AI, first behavior. So we are building like bots for sales, service, finance, and these could be like voice chat, emails, everything, and we deployed in like variety of use cases. So you are already doing like a number of pilots with dealers here and seeing like amazing, amazing results.
Kyle Mountsier:Explain the agentic, AI, just a little bit, because I think some people get lost with that, like, explain exactly what it's doing and what you're hoping to have it accomplished. Because when we talk about AI, a lot of people are talking about, like, the ability to converse with it, but you're talking about taking action. So go into like, what's like in a specific use case, what are you what are you seeing the agent, agentic. Ai, doing so.
Sanjay Varnwal:Agentic. Ai, I mean conceptual level, it is like you do set of things and that that lead to certain results happening, right? Agent A is just automating all of these things, where, instead of a human, that particular agent is doing things and then presenting you with the results that, hey, I have done this. These are the results. I mean, if you want to set auto rule that, okay, if these results are above 90% confidence, just go ahead execute them. Or I need to, like, check with it and do the action that, whether I need to take it or or reject it, right? So, Agent decay in general, is automating lot of stuff that you do. Instead of you working on the software, the software is working for you, and it is telling you the options that, okay, I've done this. Do you want to present it? Do you want to publish it? Even that card could be automated, right? So, merchandising, we are a. Doing a lot of agent again, like we did the shot, instead of users signing off, we are, we have we are. We are basically signing off all of these things. We are changing the cars images in real time on website, without informing the dealer, right, based on which image is doing better. So some of these, these things, we are, we are automating on its own, because we are controlling the website infrastructure now for media and similarly here. So some of these things, right?
Kyle Mountsier:So like, which, which image should be the first image, or something like that,
Sanjay Varnwal:one of the first use cases, right? When you get into pricing, you will be able to when you manage the ads. Let's say there is an agent, a which is, which is AD, which is contextualized or trained on the ads behavior, so it automatically will will increase or decrease the budgets or choose the mediums where it needs to spend more based on how the clicks are performing, right? So you don't need a person to do all these things for you. The agent Aki, which is trained on those use cases, will do it for you, and it is way more efficient.
Kyle Mountsier:Yeah, I think, I think that the name of the game with AI is efficiency, and it's, how do I take the tasks that can be repeated, understood when it comes to data, and hand it off to something that can do it faster and more efficiently, to free my people up to do the thing that they're best at? Sanjay, thank you so much for spending time with us today. We We walked the gamut from like, what you saw on Amazon and what we saw in the early 2000s to now, what what you're doing and what you're what you're innovating on top of and I can't wait to see what you and spine do as you deeper dive into automotive thanks for joining us today on Auto Collabs.
Sanjay Varnwal:Yeah, thanks a lot for having me here, Kyle, Paul and Mike. Great having chat with you. You guys are like huge father.
Paul J Daly:One of the things I'm super impressed with is the hyper focus of Sanjay and his team to fix small problems, and how all of those things add up to fixing big problems. Literally, that sounds maddening to me, focusing that long on the problem of reflections, but it's indicative of the fact that they understand that the small things impact the big things. And we see that a lot, even as we deploy AI, quite a bit within our business. If you don't focus on the small things, the big things end up being terrible.
Michael Cirillo:This is, like, the worst part of our human nature, which is that we want to perpetually be at the top step already. Yeah, we never want to climb the steps to get there, not realizing that every step gives us a lesson needed to actually stand at the top step. But like, you want a real world, very painful implication of why people need to focus on the smaller thing and be great at it. Look at that dude. I just watched that documentary about the guy the deep blue ocean, whatever
Paul J Daly:the oh, he went down and like that. The whole warnings
Michael Cirillo:of like, you need to focus on that thing and make that thing better. And next thing you know, it's like, implodes in the middle of the ocean, right? And, and we see that, and we're like, what an idiot. But we're doing these things in our businesses every day. We're ignoring, you know, the small thing. And so to your point about that's very inspiring to have the discipline. And
Kyle Mountsier:I'm asking chat GPT, how do I change the world today, right? It's right. How do I make sure that this employee is able to deploy this specific task and function well within my organization, those type of specific questions like, How do I make sure that mirrors show up every time I do a background replace. You know, that's how specific we have to be in everything, and especially when you're deploying AI tools across your organization. Doesn't matter if you're like doing communicative, communicative AI, agentic AI, you know, photo backgrounding, whatever it may be, being very specific with the task that that person or technology is accomplishing. Is it going to be extremely important as we move forward,
Paul J Daly:fewer better, there you go. Well, thank you to our friend Sanjay and the team at spine for putting their heart soul and all the intention to making this industry better. On behalf of Kyle Mountsier, Michael Cirillo. Michael Cirillo, new Ray Ban, meta glasses and myself. Thank you so much for joining us on Auto Collabs.
Unknown:Sign up for our free and fun to read daily email for a free shot of relevant news and automotive retail media and pop culture. You can get it now@asotu.com that's asotu.com if you love this podcast, please leave us a review and share it with a friend. Thanks again for listening. We'll see you next time.
Paul J Daly:Welcome to autocala. US,
Unknown:recording you.
