How Dresma Reached $2M ARR with 28 Customers Using Usage-Based AI Content Pricing (Interview with Co-Founder and CEO Siddharth Sinha)

Largest customer and current ARR

Nathan Latka

00:00What does your largest customer pay you today?

Siddharth Sinha

00:02>> Just about $500,000.

Nathan Latka

00:03You're at 2,000,000 right now here in December 2025. And where were you exactly one year ago?

Siddharth Sinha

00:07>> We were at about 1.3.

Nathan Latka

00:09Everybody knows Puma, the shoes. How does Puma use you guys?

Siddharth Sinha

00:12>> Puma uses us to basically localize content for all the markets that they are in. To give you an example, we recently essentially helped them launch in Southeast Asia.

Nathan Latka

00:21How low at any point in time did the bank balance get before you turn profitable?

Siddharth Sinha

00:25>> Because of tens of thousands of dollars.

Nathan Latka

00:27This year in 2025, they're finished here in December with $2,000,000 of ARR, 28 paying customers, 31 team members. If I forced you to tell me on average, what does a customer pay you per month or per year, what would that dollar range be? Hey, folks, my guest today is that dresma, an AI powered platform revolutionizing ecommerce imagery. He's a Cornell engineer and NCAD MBA alumnus as a serial entrepreneur leading his third venture as CEO and co

00:55founder of the business. Alright. Siddharth, you ready to take us to the top?

What Dresma does and how it makes money

Siddharth Sinha

00:59>> Great. Nice to meet you, Nathan.

Nathan Latka

01:01Good to meet you as well. Take us into the business. Right? So I'm gonna pull up the website here. Tell the audience what you guys How do you make money?

Siddharth Sinha

01:07>> So we basically help any brand create all the content that they need to be successful online, whether it be for marketplaces, for social media, or for their own web store. So everything from videos to lifestyle imagery to the any content that goes on their page. That's that's what we do. We help the brands do this at scale with studio quality.

Nathan Latka

01:29So if I tried to pigeonhole you here, would you say this is more of a AI like Pinterest or more of a sort of sort of network of AI images around merchandise from a specific brand or neither?

Siddharth Sinha

01:41>> So this is essentially think of us as a marketing agency on steroids. So effectively, give you all the intelligence that the agency gives you about your brand, your environment and the landscape, as well as give you the workflows to be able to create content that makes you successful in that environment.

Nathan Latka

02:01Okay. Help educate me here. Let's just dive into a specific example. So let's pick one here on your website. Everybody knows Puma, the shoes. How does Puma use you guys?

How Puma uses Dresma for market localization

Siddharth Sinha

02:09>> So Puma uses us to basically localize content for all the markets that they are in. So to give you an example, we recently essentially helped them launch in Southeast Asia. So all the content that was designed for their European market was localized for the Southeast Asian market. We gave them all the sort of intelligence as to figure out what kind of context they should be showing their products on, what kind of models they should be using

02:35>> to showcase that product, and all of it was done using AIs. They used AI models, AI locations, and essentially localized all the entire product portfolio for that market.

LLM stack and the data layer below Gemini

Nathan Latka

02:45And what sort of LLM stack are you sitting on top of? Or are you doing tool calling and you just field the request to any of the different models?

Siddharth Sinha

02:53>> So we essentially are primarily built on top of Gemini for all the content creation. We use a bunch of different video models depending on what kind of content you're creating. But for most of the imagery, we are using Gemini. And we basically sort of lay a data foundation below it so that essentially we are able to inform the the model as to the best referencing, the best prompting that can be used to create content for that

03:22>> particular brand.

Nathan Latka

03:24And Siddharth, lot of folks are saying these AI tools are just wrappers. The reason they say that is because so much of your revenue goes to paying for Gemini credits. What percent of your total revenue today goes towards paying for the credits?

Siddharth Sinha

03:36>> So it's about 25 to 30%.

Nathan Latka

03:39That's not So

Siddharth Sinha

03:41>> basically, what the brand is also paying us for essentially are two things. One is the data layer itself. So we essentially have a lot of real time intelligence as to what kind of content brands should be creating. So that's one part of the value add that they bring on top of the LLM or the data creation layer.

Nathan Latka

04:00So just to be clear, like lifestyle shots versus catalog versus marketplace versus ad, that's what you mean by the kinds of content?

Siddharth Sinha

04:07>> Yes. That's the those are the kinds of content. But we essentially you look at the sofa and the environment that that sofa is sitting in, that information comes from our data layer. So essentially, given a type of brand and a type of product, you're able to essentially contextualize the the product for the best appeal to the audience that the brand is actually pitching to.

Nathan Latka

04:28And if I'm Puma, when I logged into dresma.com, will I effectively see all of my products with basically all the related metadata and the variance on that metadata for certain geographies and product lines and human models?

Siddharth Sinha

04:40>> Yep. So basically, currently, we are integrated with with Amazon seller panel as well as a couple of different pins so that you can actually pull pull your data directly from from those sources. And we essentially have built a profile of your brand and are able to track what kind of content is working for brands like yours in different markets. So then we use that intelligence to be able to to people to better advise the brand as

05:09>> to what kind of content they should be creating.

Usage-based pricing model explained

Nathan Latka

05:11Let's shift to pricing. I either see people charging $5 a month and they want a million customers, or they don't even show their pricing and they want 10 customers paying 5,000,000 per year. You do both. You've got $5 over here and enterprise, let's talk over here. What's the model?

Siddharth Sinha

05:26>> Mhmm. So essentially, we we are on a usage fee basis. So depending on how much content you're creating, that's that's how much we charge you. And essentially, the the customization of the platform comes with usage. So as as you're actually coming in, signing up, and and creating content, that's how we learn more about your brand and what kind of content works for you. And then the the customizes by itself, so there's no cost to the usage

05:50>> of the platform itself. Essentially, just just the content that you're creating that's that's so basically, the usage fee for that.

Nathan Latka

05:58Yes or no. Do you upsell based off number of seats from the Puma marketing team?

Siddharth Sinha

06:04>> No. So so essentially, it's just the quantum of content that you create.

Nathan Latka

06:08Okay. And last question here, true or false, you don't charge for product based upselling, just usage?

Siddharth Sinha

06:14>> Yes. Just usage.

Nathan Latka

06:16Okay. Very cool. So very focused on usage. You can see sort of how those credits work right here. Siddharth, I'm gonna force you into an average here, which I know engineers never love, so bear with me. Okay? If I forced you to tell me on average, what does a customer pay you per month or per year, what would that dollar range be?

Average ACV and credit pricing mechanics

Siddharth Sinha

06:34>> For a year, it'll be about 60 k.

Nathan Latka

06:37Okay. You answered that fairly quickly. What does that mean in terms of total credit consumption for that 60 k?

Siddharth Sinha

06:43>> So 60 k essentially in platform is 60,000 credits. It's the dollar credit pretty much One to one. On that. Yeah. It's one to one. Only do you when you get into very high high quantum consumption, so upwards of say $200,000-$300,000 do we discount on the credit pricing.

Nathan Latka

07:02Okay. Okay. That makes a ton of sense. And give us the backstory. Now that we understand the pricing and the product, what year did you launch the business?

Company founding story and image processing origins

Siddharth Sinha

07:09>> So the business launched in the summer of twenty twenty. So we started off with essentially automating all the workflows that that happen after photographs get shot in a studio. So essentially started off there. So we built up built a whole bunch of image processing algorithms that automated essentially what people would be doing using Photoshop. So that's where we started. So we have a very strong core expertise in image processing and understand what kind of images work

07:41>> for brands. And then over the last year and a half, two years, obviously, we've sort of adopted all the new AI tools and essentially built that whole creative imagery layer on top of the image processing expertise that we have.

Nathan Latka

07:53And how much did you spend to launch the business?

Siddharth Sinha

07:57>> From a tech investment. So I mean, in terms of total capital raised so far, we raised $3,000,000 and we are about a revenue of just under 2,000,000 right now.

Nathan Latka

08:09Oh, that's great. Okay. So you'll do you're at 2,000,000 right now here in December 2025, where were you exactly one year ago?

Siddharth Sinha

08:15>> We were at about 1.3.

Nathan Latka

08:18One point are you happy with that growth?

Siddharth Sinha

08:20>> Not really. I would have liked a little bit more than that. But what we've seen, I mean, I think with a lot of different AI or generative AI companies is that there's been a lot of sort of POC work this year, not a lot of, you know, actual production. So that's changed in the last couple of months. We were actually seeing a lot of brands now adopting AI generated content for their bulk workflows for their actually,

08:49>> core the core business as well.

Nathan Latka

08:51Guys, remember, I am not just a YouTuber. I'm investing into my third fund. We've deployed $250,000,000 into 550 software companies so far, again, at founderpath.com. If you're interested in capital, I would love to cut you a check because I know you're investing in your education. You watch my show. So sign up at founderpath.com, and when you get the onboarding email, I reply and I see all those. Just reply and say, Nathan, I found you through YouTube,

09:14and I'll make sure to prioritize you. I would love to cut you a check. Check out founderpath.com.

Siddharth Sinha

09:19>> Mhmm.

Nathan Latka

09:20And so you're doing 2,000,000 of revenue right now. How many total customers are paying?

Siddharth Sinha

09:25>> We have about 27 or 28 different customers.

Studio partner go-to-market and Champion example

Nathan Latka

09:31Okay. And don't name them obviously, But what does your largest customer pay you today?

Siddharth Sinha

09:37>> Just about $500,000.

Nathan Latka

09:39Okay. That's that's pretty impressive. Right? To be at 2,000,000 of revenue, you've now proven that there is a usage based upsell enterprise model where you can expand those accounts to 500,000 per year. That must get you excited.

Siddharth Sinha

09:52>> Yep. Yep. So we have two kinds of customers, Nathan. So we work directly with the brands, but, we also do work with aggregators. So we were we have partnerships with studios where we co sell a solution to brands. So a lot of brands still need the product photography done. So they get to get the studio to actually do the product photography and then they use our AI algorithms to then transform those images into model imagery, lifestyle

10:18>> imagery, videos, and all sorts of different workflows after that.

Nathan Latka

10:21Give me the URL of an example studio you work with. That's one of your good partners.

Siddharth Sinha

10:26>> So we work with ePhoto in Italy. So ePhoto works with most of the large luxury brands in Europe. So they do photography for them and they're using our

Nathan Latka

10:40Is this is this it? This is their studio, something like this?

Siddharth Sinha

10:43>> Yep. Yep.

Nathan Latka

10:44Is this their website?

Siddharth Sinha

10:45>> Yes. That's it.

Nathan Latka

10:48Okay. So now that my audience is seeing this here in the in their YouTube recording here, explain your go to market again and how you work with them.

Siddharth Sinha

10:55>> So we work So essentially, if you look at somebody like, for example, we did work with Champion recently, there's actually an ongoing project. So Champion, essentially what ePhoto does is essentially does flat lay shoots of all their products. They use our workflows then to take those flat lay shoots and convert them into model images, into videos, into lifestyle, location imagery. All of that is done using our AI workflows. The initial photography is done by ePhoto.

Nathan Latka

11:24I see. I see. I see.

Siddharth Sinha

11:25>> It's almost like when we're building synthetic models to train to train the foundation models, basically, you are the data input. You are the, you know, they are sorry. They are the raw input.

11:35>> Right? Are the raw input.

Nathan Latka

11:37Yeah. You're then generating synthetic data on top.

Siddharth Sinha

11:41>> Exactly. So if you look at the if you look at what we have where AI can really help, I mean, you know, the product still is unique to the brand. It has very unique characteristics. So the AI can't really generate the original product itself. So that really needs still a good high quality image to get started with. Then you can take that image and then adapt that into all sorts of different outcomes.

Revenue breakdown by channel

Nathan Latka

12:04So let me ask you a question. How much of your revenue growth would you say has come from this partnering with studio strategy?

Siddharth Sinha

12:11>> So currently about to I think 50% or 55% of our revenue comes from that channel. We have studio partners in Western Europe, in Germany, in France, in Italy, as well as in a couple of them in The US as well. So we are geographically diverse when it comes to partnerships that we've established in all our major markets.

Nathan Latka

12:36So true or false, you've scaled from 1,000,000 to 2,000,000 of revenue over the past twelve to eighteen months. 50% of that growth has come from partnerships.

Siddharth Sinha

12:44>> Yep. Yeah.

Free tools, programmatic SEO, and traffic

Nathan Latka

12:45Okay. Cool. Any other growth channels you're using to scale? I notice in your footer, you're investing in these tools, and these can sometimes do really well for SEO and LM, sort of programmatic SEO optimization. When we go look at your actual traffic hitting your website, that free tool strategy brings in about 4,300 clicks per month. Is that intentional?

Siddharth Sinha

13:05>> Yes. So we do use so essentially, for that, that's really a lead generation channel for us. Effectively, that gets people to come in, try out the technology, make sure that, you know, the the the core AI is there. And then we work with the with the branch essentially to upscale upsell and essentially get the get the platform adopted.

Nathan Latka

13:26Yep. This is a huge Guys, all of watching, you should copy what Siddhartha's done for your own industry or niche, right, which is make these free tools with very little friction in your footer. It's an SEO play. Invest in the SEO all the way down the page so it pulls in free traffic. Right? And these are basically the small components, Siddhartha, of what your larger product does once they actually start paying.

Siddharth Sinha

13:47>> Exactly.

Nathan Latka

13:48This works well for you. Your top performing one is your HD photo converter that's representing almost 50% of that total traffic. A lot of people, so they'll say, yeah, Nathan, but this traffic doesn't convert. How would you respond to that? You're seeing the data.

Siddharth Sinha

14:01>> So essentially, what that does is it allows us to a I mean, get the allow the customer to have a a small foot in our door. Right? So it's just it gets us gets them onto a route or to our website, gets us aware gets them aware of the solution that we are offering. Essentially, there these are typically the way we sort of choose these these keywords are these are typically sort of small wins that that

14:25>> a marketing team or or the studio team is looking for. So they may have a bunch of images that are of the wrong format, and they need them changed, they need that need them done quickly. So just this gives them a quick salute, quick win. So essentially come on to our platform and this socialize allows us to socialize the solution with them.

ABM tech stack: SmartLead, Apollo, and Clay

Nathan Latka

14:43Mhmm. Are there any other growth channels you use to add over a million dollars of revenue over the past, you know, twelve to twenty four months?

Siddharth Sinha

14:50>> So we've essentially have done quite a bit of outbound work over the last year as well. So essentially a mix of email and LinkedIn outreaches. So we essentially have an orchestration platform that we built using some of the data enrichment tools out there to be able to essentially reach out to these to to our target audience in a very specific way.

Nathan Latka

15:18So explain that tech stack to me with this graphic, Siddharth. This is a graphic of the 33 growth tactics CEOs tell me over and over. You're talking about account based marketing here. Already educated us on how you're doing programmatic SEO with free tools and you already educated us on how you're working with sort of these value added resellers or partners with the studios. Mhmm. You're also using ABM. So teach us here. You've just broken $2,000,000

15:40of revenue. What does your ABM stack look like? What technologies and tools are you using here?

Siddharth Sinha

15:46>> So, basically, essentially for the actual reach out and sequencing, we're using smart leads. We pair that with a sort of an add on sort of orchestration tool that we sort of built built in house that helps us do sort of LinkedIn automation as well. For the data, we use Apollo and Clay as enrichment tools. So essentially, Apollo as the data source and then Clay is the enrichment enrichment tool.

Nathan Latka

16:13So this site, SmartLead. Right?

Siddharth Sinha

16:16>> Mhmm.

Nathan Latka

16:17Plus Apollo, plus Clay, plus some custom stuff you built in house. Yeah. That is your ABM strategy.

Siddharth Sinha

16:22>> That's true. Mhmm.

Nathan Latka

16:24Of the $1,000,000 of new revenue you added in the past twelve months, how much has come from ABM?

Siddharth Sinha

16:29>> That would be about, I would say, 30 to 35%.

Team size, sales structure, and AE quotas

Nathan Latka

16:35That's a big chunk. What's your team size today and how many are dedicated to sales and marketing and outreach?

Siddharth Sinha

16:42>> So total employee count at the moment is 31. In terms of sales and marketing, we have five people right now across sales and marketing.

Nathan Latka

16:54And for folks that are also around $2,000,000 of revenue looking to scale and they're watching this, they wanna say, well, wait, Nathan. How did Siddharth structure those five jobs? What are they? Are they all SDRs, AEs? What's the mix? Who carries a quota?

Siddharth Sinha

17:07>> So essentially, essentially, currently they're two AEs. We don't have SDRs at the moment. We have two AEs and we have three people that are actually sort of two and a half people looking at marketing right now.

Nathan Latka

17:22Okay. So three marketing, two AEs. When you hire your first AE and you're a seasoned entrepreneur, you've sold to other companies, right? What what quota do you give that AE when they start?

Siddharth Sinha

17:32>> So we are looking at starting out with an AE in the first year, we would look at spring somewhere between 500,000 to $750,000 worth of revenue.

Nathan Latka

17:46Okay. That's their quota target. And generally speaking, and don't reveal actual salaries and names, but generally speaking for an AE when you hire them at your stage, what's the base you're gonna pay them?

Siddharth Sinha

17:55>> So currently, actually, AEs are based in India. So we've actually had a very sort of capital efficient that way, so our costs are actually very low. So we're essentially selling out of almost the entire team sits out of India right now. All of our tech development happens out of India, so we actually have a very low cost base when it comes to.

Nathan Latka

18:18Those two in free marketing. Those five on sales and marketing in India, which city, Chennai, Bangalore, somewhere in Mumbai?

Siddharth Sinha

18:25>> We're in near Delhi in Gurgaon.

Nathan Latka

18:28Okay. Delhi. Interesting. So are they are they totally commission based, or do they have any base salary at all?

Siddharth Sinha

18:34>> No. They have base salary plus commission.

Nathan Latka

18:36Okay. So base, is it fair to say it'd be under $20,000 per year, 20,000 USD?

Siddharth Sinha

18:42>> The average of the two would be 20.

Nathan Latka

18:44Okay. So, guys, there you have it. There's another good data point for you. Interesting. Are those people saying, Nathan, I can't have AEs in India selling to big Puma companies based in California. Is that true or false based off your experience?

Siddharth Sinha

18:57>> That's false. So you can have people based in India sell to large enterprise US customers. You do need some face time. So myself,

19:07>> my co founder, and the sales person at times will be traveling to The US and Europe. So those are seasonal or periodic trips that we make to make sure that we are in front of our main customers as often as we need to.

Nathan Latka

19:22Mhmm. And Siddharth, so as we sort of wrap up here, wanna spend one or two minutes on fundraising questions and one or two minutes on where you see your product and your space going before we wrap. So on the fundraising, you raised 3,000,000. When was that round done? Was it all one round?

Seed round, valuation, and fundraising history

Siddharth Sinha

19:36>> This was a seed round. That was the the only external capital we raised that was done at the end of twenty twenty one.

Nathan Latka

19:44Okay. So 3,000,000 seed. And was that a to traditional safe?

Siddharth Sinha

19:49>> No. That that was a price round.

Nathan Latka

19:51It was priced. Okay. Look. I have to ask this. What was the valuation?

Siddharth Sinha

19:56>> At that time, we did a post money of 10,000,000.

Nathan Latka

19:59Okay. How did you how did you defend that? People will say, wait, it's pre revenue. There's not even one customer yet. How is he getting a 10,000,000 valuation? At

Siddharth Sinha

20:06>> that time, we did have revenue. So we were at about a book revenue of about 400 or $4.50 ks at the time.

Nathan Latka

20:13Okay. So you use that traction to help defend Did post you rely on your past companies that you were at as well? You had some cache there, some experience?

Siddharth Sinha

20:22>> Some experience there. Plus, I mean, both my co founders have very deep expertise in e commerce, especially content for e commerce. So they come in with this. So basically, they used to run studios for large e commerce enterprises in India, as well as doing post production work for large studios in Europe. So they come with deep expertise in the vertical.

Nathan Latka

20:47And are you default alive today? In other words, are you profitable?

Path to profitability and lowest bank balance

Siddharth Sinha

20:51>> So last quarter was our first profitable quarter. So we are profitable.

Nathan Latka

20:58Congrats. That's exciting. How low at any point in time did the bank balance get before you turn profitable?

Siddharth Sinha

21:04>> Because of tens of thousands of dollars.

Nathan Latka

21:08How did that feel? Were you nervous?

Siddharth Sinha

21:10>> It's very nervous, but I mean, I've run bootstrapped businesses before, so I know how it goes. It's okay. Obviously, as they say, cash is more important than your mother, so it's it's important to keep keep an eye on it, but, you know, you you have to stay a little bit conservative in those times. But, you know, we were very confident of of the business and the value we were bringing to our customers. So we didn't really

21:41>> have I mean, I wasn't that nervous.

Nathan Latka

21:44Yep. Very good. Well, hey. Take us home here. We're on dresma.com right now. The headline says your visual AI content creation engine. You know what products you're building for the next year. If I refresh this a year or two years from now, what's your headline gonna say?

Product vision and replacing digital marketing agencies

Siddharth Sinha

21:58>> It's gonna be I mean, essentially, we'll be we'll be substituting digital marketing agencies. You wouldn't need digital marketing agencies anymore. You'll essentially be coming to us and we would essentially create your entire content plan, whether it be for whichever content channel you need to reach your customers more.

Nathan Latka

22:16Guys, there you have it from Siddharth. Dresma launched back in 2020. Did his 3,000,000 pre seed at ten million post with $400,000 of revenue back in 2021. Fast forward to 2024, they broke $1,300,000 of revenue. And this year in 2025, they're finished here in December with $2,000,000 of ARR, 28 paying customers, thirty one thirty one team members, profitable in q four, which we love. And get this, they've upsold against their usage based pricing model where their

22:45largest customer today is paying $500,000 per year. He's on to something, but it wasn't all easy. His lowest bank balance as he was building was in the, quote, tens of thousands before he turned profitable. Now reinventing the visual AI content creation space specifically for ecommerce brands that need to put the same shoe in six different languages with seven different models on 1,000 different websites. Check it out at dresma.com. Sidar, thanks for taking us to the top.

Siddharth Sinha

23:11>> Thank you, Nathan. Good talking to you.