How Akool Hit a $36M Run Rate and 30,000 Customers in 2026 (Interview with CEO Jeff Lu)

Introduction and Guest Background

Nathan Latka

0:01Hey folks, my guest today is Jeff Liu. He's the founder and CEO of Acool, a generative AI video platform for avatars, video translation, and real-time video. He's an AI and computer vision veteran with over 10 years of experience in this space, having previously worked at Microsoft, Apple, specifically on Face ID, and Google Cloud. Jeff, you ready to take us to the top?

Jeff Lu

0:23Yeah, thank you for inviting me here today.

Leaving Google Cloud to Found Akool

Nathan Latka

0:25You bet okay, what year did you leave Google Cloud or Apple to launch a cool?

Jeff Lu

0:30Yeah, so I left ⁓ Google Cloud in two thousand twenty two to launch a quote, so it's kind of at the beginning of this generative AI wave.

Nathan Latka

0:42Mm-hmm. And what were you seeing at Google Cloud at the time that sort of gave you the idea for this? What was the initial product you launched?

Spotting the Generative AI Opportunity

Jeff Lu

0:49Right, so ⁓ I have been working on their generative AI for images and videos for a pretty long time. And ⁓ what we see is that the technology is becoming ⁓ mature and ⁓ but Google is not working on it. At that time Google is pretty conservative on like ⁓ generative AI, especially on videos and so on. They they think they cannot control the results. And they want everything to be one hundred percent under control. So ⁓ I I I I've been working on it for a very long time and I see a technology breakthrough and I see lots of adaptations and things coming out and so on. So ⁓ I decided to full time quit a job at Google and ⁓ start a company. But before I was kind of ⁓ part time, work a little bit and it's during the COVID and so on, yeah.

Revenue Transparency and 2025 Numbers

Nathan Latka

1:44No, Jeff, I wanna keep getting your backstory and see how you got your first a hundred customers and how you grew and how you're building product and you built your own foundation model. I wanna dive into that. I don't wanna bury the lead though. You've been pretty transparent about your revenue growth. Are you comfortable sharing what revenue is today?

Jeff Lu

1:59Yeah, yeah. So ⁓ for twenty ninety five our ⁓ recognized revenue is about twenty five million dollars and we are ⁓ more than double this year. So we are expecting to get ⁓ sixty or something. Yeah.

Nathan Latka

2:16So you anticipate finishing in December twenty twenty six with about sixty million of ARR?

Jeff Lu

2:22⁓ revenue, not AR. So ⁓ yeah. Yeah, so revenue you recognize the revenues and add ⁓ the trailing twelve months. So that's how it works. And the AR you probably use the ⁓ last month's multiply twelve, that's ⁓ revenue run rate. So it's a different ⁓ way of measuring it. So yeah.

Nathan Latka

2:25Tell me how you measure. So you're measuring the more conservative way, you're adding up trailing twelve months revenue.

Jeff Lu

2:50Yeah, yeah, yeah. That's that's how we ⁓ do it now. So I know I know the market, people trying to do it ⁓ in various more aggressive ways, but we kind of ⁓ intend to do it more connovative way. So yeah.

Nathan Latka

3:04Jeff, I love that about you, right? That's the quickest way to tell if I'm interviewing a marketing founder or a technical founder. All the marketing founders take last month's revenue, multiply by 12. You're conservative, you're looking back, which I think is awesome. What does that mean in terms of we're recording this in July? If we just look at last month's revenue, how much did you do just last month?

How Akool Measures Revenue

Jeff Lu

3:25Yeah, yeah. So ⁓ we last most revenue I think it's over three million dollars. Yeah.

Nathan Latka

3:31Wow, that's incredible. How d what was that like? I mean, did you ever expect when you launched this thing three, four years ago, you'd be doing three million dollars a month?

Jeff Lu

3:39⁓ I think ⁓ I think I haven't thought about it and especially ⁓ it's actually pretty different when we gr just get started. We're trying to focus on B2B purely and we cannot get into mix of B2B and B2C and and so on. So but our goal is still to bring this to a very big company. So ⁓ I think we are still on the way ⁓ going there.

Nathan Latka

4:06Okay. Well come on. I wanna see your product. There's so many people building AI tools. They say it's the best. I asked you ahead of time, prepare a screen so you can share and show us the product. Show us what you're building.

Jeff Lu

4:17Sure, sure. Let me let me share the screen and share the product.

Nathan Latka

4:24There should be right on Riverside at the bottom a little share button you can use. There we go. Okay, so let get help us get context first, Jeff. So what are we looking at? Is this a logged in customer experience?

Jeff Lu

4:28Yeah, ⁓ Yeah. So this is the logged in customer experience. So after you log into our platform, you will see the dashboard and so on.

Nathan Latka

4:42Okay, this is great. So take us through your most popular tool. What's the number one feature people use?

Product Demo and Core Features

Jeff Lu

4:44And Yeah, so ⁓ we have several tools pretty popular. So the thing that we initially get popular is the swap features. So we have a whole swap series. You can swap the face, swap the character in the video, swap the motions, swap the head, and swap the hair, and so on. So the swap series is the initial thing that's ⁓ very popular. And ⁓ then we get our avatar features are very popular, especially around the on the B2B side. Our most popular feature is actually avatars. And you can create static avatars, and you can create live interaction avatars, you can create your own version of the avatars, and so on. This is the most popular feature on the B2B side. And on the video foundation model, we are also very popular as well, and we train

Nathan Latka

5:40Very interesting. And why so people look at the screen you're showing right now and they're going, Wow, it's a bunch of like individual tools. Oops, sorry, Jeff, the connection's a little laggy. So I'll I'll I'll pause there. You keep going. What else do people use?

Jeff Lu

5:46⁓ ne swap Yeah, yeah. And ⁓ the foundation models which ⁓ we ⁓ do some fine tunes are also very popular. So we trained the Avatar models ourselves in-house and a swap series ⁓ mainly by ourselves and fine tune the foundation models ourselves and more recently we launched ⁓ something called Agentic Canvas and it's a new feature. And I think ⁓ it's also getting popular. ⁓ basically you can use AI agents to help you to create a long format of the videos. ⁓ and you can just ⁓ type and it will create the video by itself.

Nathan Latka

6:35Back, Jeff.

Jeff Lu

6:36Yeah, yeah, I'm back.

Nathan Latka

6:38Do you have internet you can switch to just to make sure we keep this really high quality? I wanna make sure you look good.

Jeff Lu

6:43Okay, so it's slacking.

Nathan Latka

6:45Well now it's fine, but you just cut out for a second. Is it good? Is it stable now?

Jeff Lu

6:49⁓ y yeah, it seems when I using the website and sharing screen it's a little bit slow, so

Nathan Latka

6:54Screen share. Okay, so we won't screen share. Let's pick this back up. You mentioned you're fine tuning the models. My question is a lot of people are gonna go, Jeff's not doing anything special. There's a bunch of tools I can use to do a face swap, but you've actually you're not just using, you know, other foundation models. You're actually fine tuning. Now you're a technologist. Explain to the marketer watching this why that's valuable to them, why your fine tuning adds value to their life.

Jeff Lu

7:18Yeah. Yeah, yeah. For the swap series of the models, we're actually not just fine-tuning. We are designing the model from scratch and ⁓ doing the models ourselves. ⁓ so it's ⁓ pretty different model architecture and ⁓ training data size and so on. So the result quality and how it handles ⁓ corner cases and boundary cases are very different. ⁓ and then for the fine-tune it happens more to the ⁓ image to video models and ⁓ so on.

Nathan Latka

7:50Can you just educate us? When you say you're building your own foundation models, most people think you have to have billions of dollars to do that. How are you doing it?

Jeff Lu

7:59Right, right. So ⁓ we are not building the largest video foundation models. We are building like ⁓ video models for specific tasks. So for example, the avatar, you generate humans in the video. Human generation for only that task and then the model is much smaller. And also we create swap series, you make changes to humans in the videos, and these models are much smaller. than the large foundation models that generate everything. So that's how we control the cost and control the budget and make things move.

Nathan Latka

8:34What is your actual tech stack look like? I mean, do you have a bunch of Mac studios running in a closet or H two hundreds somewhere? Are renting from one of the inference providers? Like, how do you do your training?

Jeff Lu

8:44Yeah, yeah, we launch ⁓ we launch servers from their ⁓ cloud providers and then we do training on these servers.

Nathan Latka

8:54I see. Who do you use? Are you using together or hyperbolic? Or which company, which inference company are you using?

Jeff Lu

9:01We use multiple ones. but Lambda is used a lot and there's some others as well depend on the workload and the needs and so on.

Nathan Latka

9:09Since you're a user of those platforms, I'm trying to figure out who's gonna win that space. Who do you think is gonna win that space?

Jeff Lu

9:16Yeah, I think on the infrastructure side there will be multiple winners. It's ⁓ it's very hard for a company to dominate. So that's how the B2B side works. When you are going to B2B, then ⁓ I think there will be ⁓ multiple player game. And if you are doing B2C then ⁓ probably there will be ⁓ a a bigger player, so yeah.

Nathan Latka

9:42Other folks listening right now are trying to like build their own version of sort of foundation models. And I think the tr my audience generally, they know what a prompt in Claude is, right? Or ChatGPT. And then they say, Well, I want my prompt to have memory. So then they upload a bunch of MD files, but then they ha hit context window issues and hallucinations in Claude and ChatGPT. So then they say, Maybe I should host my own files, do my own training. And then it gets really complex really, really quickly. You've obviously pushed that pretty far the other direction. Do you have

Jeff Lu

9:56Yeah.

Nathan Latka

10:09Do you have these same issues like context window issues and hallucination issues when you're building your own foundation models for specific use cases?

Jeff Lu

10:16in the video space, I think ⁓ that is ⁓ a little bit less concern. The video space, the main thing and how to make it more realistic. ⁓ and ⁓ making it realistic is the biggest challenge and also make it accurate to the prompt, follow the prompt and the realistic are probably the hardest challenges. ⁓ and in our case, since we are doing more specialized ⁓ video models. And ⁓ we generally don't experience ⁓ vive issues.

Nathan Latka

10:51Jeff, can you give me can you give me your most clever example of a sentence or two that you had to put in your fine-tuning video agent that generated a really good output that you didn't think would generate a really good output when you added it?

Building Proprietary Foundation Models

Jeff Lu

11:05yeah. So actually one of the interesting things is I begin to ask the video agent to generate ⁓ like ⁓ videos for some famous books, like for each chapter and so on, and asked it to generate ⁓ a video for it. And the result is pretty amazing. It's kind of ⁓ be much better than I expected. ⁓ but ⁓ When you look at the detail, of course, there's a lot of the things that need to be updated and ⁓ so on. I think one of the one of the difficult things over there is create consistent voice across the whole video and ⁓ make it look good and so on. And yeah, the on on the on the voice voice side when you're trying to create a drama and ⁓ so on, that's probably harder. But ⁓ in general, I think using agents to create stories already ⁓ Pretty good, but still have lots of space to improve.

Nathan Latka

12:05I published my book in twenty nineteen. Sixty thousand words. We sold a lot of copies. Washington bestseller list. How close am I to be able to use a cool to update, upload the PDF of my book, plus maybe the audio book? So you have my voice, plus an image or two of me, and you make it a video.

Jeff Lu

12:16Yeah. Okay. that depends on what kind of video you want to make. If it's an avatar video, talking about reading the books, that's pretty simple. I think we can do it very easily. And if you are talking about the it's like a story and you want to see a scenery of the what the book tells, I think I think we can already do quite many of these works piece by piece and should not be far to get the whole book to work. Yeah.

Nathan Latka

12:47Interesting. I love this. Okay. what do you I guess let's get more of the growth story. When you launched this business back in twenty twenty two, how did you get the first a hundred customers?

Jeff Lu

12:58Yeah, yeah. So it's ⁓ when we first launched, ⁓ we try to close some B2B customers and ⁓ some larger ones and the process is very slow and there are lots of users come using it by itself. So we just put a payment gateway on it and then people begin to pay. So it's kind of we get some organic traffic and people come by themselves and just pay by themselves online.

Tech Stack and Inference Providers

Nathan Latka

13:28But really try to remember, I mean, when you launched, you didn't have traffic. Now today you're getting two hundred ten thousand organic clicks a month based off AHREF's data that I'm showing here on my screen right now. How did you get the traffic in the early days? How'd people know about you?

Jeff Lu

13:43Yeah, yeah. ⁓ in the early days, ⁓ we mainly get the data from the ⁓ get the users from the community. So quite many communities like ⁓ Rabbit and ⁓ Facebook groups and ⁓ Discord and so on and we kind of show lots of the product demos, lot without demos and we show quite ⁓ interesting and good looking ⁓ without demos and the people come to our And then they spread w what about.

Nathan Latka

14:16So I see you have your own Reddit channel, R forward slash O'Cool underscore official. Has this was this the major driver of your first a hundred customers or no?

Jeff Lu

14:25No, this this Reddit ⁓ this subreddit was created much more recent. It's probably ⁓ a year ago, but ⁓ it was not from this subreddit.

Nathan Latka

14:37Mm-hmm. So you are just posting in random other subreddits?

Jeff Lu

14:42in related ⁓ communities for showing demo without demos and so on.

Nathan Latka

14:48I see. Okay. What else did you do to get your first a hundred customers? So Reddit community is what else?

Jeff Lu

14:54Yeah, yeah, I think there are some ⁓ builder groups already using video creation tools and we try to ⁓ talk with the customers and ⁓ frequently schedule like one-on-one meetings and ⁓ listen what they need and recommend them tools and so on. That's that's how we initially get started.

Nathan Latka

15:18Can you name one or two of those communities?

Jeff Lu

15:21Yeah, I think there are lots of communities on the ⁓ Facebook and ⁓ Discord and ⁓ just just to find who is very active and ⁓ ping them and talk with them and recommend them. And also on Fiverr. So Fiverr has a lot of their people ⁓ using helping others to create videos and ⁓ we also talk with the people on the Fiverr and ⁓ recommend them some tools and so on.

Nathan Latka

15:48Did you create your own Fiverr account to do avatars for people using Fiverr to get that video work done?

Jeff Lu

15:55no, no, not really. We just we just ⁓ we just talk with video creators on the file.

Nathan Latka

16:01So you told them, Hey, look, you're on Fiverr giving people video services like face swap, you should use a cool to get that work done faster.

Jeff Lu

16:09⁓ yeah, we d we did that work ⁓ as well. So we did that before.

Nathan Latka

16:14I see. Okay. And build up the revenue history for me. So first line of c did you write the first line of code in twenty twenty? Are my notes right?

Jeff Lu

16:23s yeah, about right, twenty twenty one or twenty twenty or something like that, yeah.

Nathan Latka

16:28Okay, what did you finish twenty twenty one at with total revenue?

Jeff Lu

16:31It's very small. So ⁓ I think ⁓ at that time I was part time working on it. So it's a number is very small, around a hundred K or something.

Nathan Latka

16:42Okay. And in twenty twenty two, what did you finish with?

Jeff Lu

16:46W also about the Ham J both years. Yeah, yeah, yeah. Pretty flat for there. Because the first first turn one is a part time and turn and two is ⁓ is ⁓ I just get out. I think ⁓ at that time we are we spend the most of the time pursuing large contracts, but ⁓ was ⁓ not ⁓ having having having difficulties working

Nathan Latka

16:48So pretty flat.

Jeff Lu

17:13clothing type of the large contracts and it y it didn't grow much. So

Nathan Latka

17:17And what did you finish with in twenty twenty three?

Jeff Lu

17:19Twenty minutes three is two to three million dollars.

Nathan Latka

17:23Okay. And my notes tell me that that was your first big enterprise account. You landed Coca Cola and they ran a six month campaign swapping several million faces. Tell me how you closed that deal.

Jeff Lu

17:34Yeah, that's that's inbound. So that's inbound and they fund us and they ⁓ they see their ⁓ our technology the best on the market and ⁓ so they benchmark it a lot, so ⁓ they decided to work with us.

Nathan Latka

17:51So what do you do you have AEs on your team that are looking for all the prosumers that sign up for a cool for fifty bucks a month? And if you see someone at Coca Cola dot com, you proactively reach out to them or how how does that actually work?

Jeff Lu

18:04For this case at that time we don't have and it's ⁓ fully they reached out to us and they even didn't disclose their identity at the beginning until like ⁓ from the first conversation to get engaged is about three months in the middle. So ⁓ they they spent three months benchmark testing different stuff and the designing and so on. ⁓ yeah. So but ⁓ we get our first A account AE in twenty ninety four and then ⁓ begin to do more of their sales and the management work and so on.

Nathan Latka

18:39And so what did you finish twenty twenty four with in terms revenue?

Jeff Lu

18:43Yeah, twenty ninety four we gap revenue around ⁓ twelve, thirteen million dollar.

Nathan Latka

18:51Okay. And have you done all this? We know what happened after that, 'cause you said in twenty twenty five you finished with about twenty five million of revenue and you're gonna more than double that here in twenty twenty six, right?

Jeff Lu

19:01Yeah, yeah, yeah, yeah, that's right. So, ⁓ yeah.

Nathan Latka

19:05Okay. So tw just to convert twenty twenty five total revenue was twenty five million.

Jeff Lu

19:10⁓ yeah.

Nathan Latka

19:11Okay, very cool. ⁓ walk me through the team size today. How have you structured it and how many folks are full time?

Jeff Lu

19:19Yeah, yeah. So ⁓ our team is pretty dynamic. It it depends on their needs and so on. So we have ⁓ sixty to seventy people ⁓ on board and ⁓ majority of them are engineers ⁓ and ⁓ some go to market marketing sales and so on.

Nathan Latka

19:43How many are engineers? Like forty?

Jeff Lu

19:46Yeah, around that.

Nathan Latka

19:47Okay. And do you have any quota carrying sales reps or no?

Jeff Lu

19:51we have. We have.

Nathan Latka

19:53How many and how did you decide what quarter to give them when they joined?

Jeff Lu

19:56Yeah, so we have about four or five quota curling sales reps and ⁓ it's pretty standard to give one million dollar quota per person. ⁓ and ⁓ yeah, but the sales ⁓ performance usually diverse and ⁓ the high performance ones can do really well, so yeah.

Nathan Latka

20:18I'm curious how that all works. Maybe let's reverse engineer it. So so I guess how many total customers are you serving today? And what's the average customer paying per month?

Jeff Lu

20:27Yeah, we have differ we we serve across different ⁓ customer profiles, including like ⁓ enterprise, S and B, prosumer and so on. On the prosumer side we have a lot. We have tens of thousands ⁓ of customers. And ⁓ on the enterprise and S and B side we have several hundred. So that's that's how it works. Yeah.

Early Customer Acquisition Strategy

Nathan Latka

20:53So is it fair to say all of your cust like paying customers added up together you have more than thirty thousand paying?

Jeff Lu

20:59I currently pay, I think I think so. Like twenty to thirty thousand, around thirty thousand maybe.

Nathan Latka

21:06Interesting. But most of your growth is coming from the enterprise sales.

Jeff Lu

21:10more recently, yeah. More recently more growth come from the enterprise sales.

Nathan Latka

21:14I think people watching right now are gonna be wondering how big like how much value can you add to one enterprise account if they start off paying fifty bucks a month? Are you comfortable? Don't share the name, but are you comfortable sharing your largest contract value today?

Jeff Lu

21:29Yeah, yeah. So currently our largest customers pay us it's it's actually usage based, but they pay us like around ⁓ four million dollars a year. And

Nathan Latka

21:44What you have individual customers paying four million a year.

Jeff Lu

21:49No, no, no. They are contract on the B to B side. ⁓ an individual self service at minimum they pay us ⁓ twelve dollars a month, right? So

Nathan Latka

21:51Yeah. No, but just to be clear, your largest enterprise accounts today, you have some individual logos or brands paying you more than four million per year.

Jeff Lu

22:06That's that's our largest account, not individuals.

Nathan Latka

22:10That's still pretty incredible. What, what, what ⁓ makes them scale so quickly? When you say usage-based, is it number of credits? How do you what how what value do you assign to each credit? How do they consume the credits?

Jeff Lu

22:22Right, right. So ⁓ we we provide ⁓ credits and to the users that's ⁓ using their platform and it's usage based and some of them actually integrate into their software and ⁓ when they scale up their usage, ⁓ they also scale up the API course. It's all through API usage.

Nathan Latka

22:44Yeah. Jeff, I have to tell you this respectfully. I asked this question because when I look at your pricing page, I mean this thing looks like shit. I'm so confused. I'm like, where do I go? There's so many sliders. What do I pay for? But this is working, huh?

Jeff Lu

22:58⁓ so yeah I think we we we provide one more tier. So before we only provide pro, pro max and business. And ⁓ we just recently added a starter plan. So we kind of ⁓ give people more options to get started.

Nathan Latka

23:18But do you I guess what I'm asking is like when I land on this, I'm like, my gosh, I'm so confused. Like, what do I get for twelve thousand credits? Like, is that three face swaps or a thousand face swaps? How do you make your new users quickly understand what they're paying for?

Revenue History from 2021 to 2026

Jeff Lu

23:32⁓ I think if you scroll down there is some explanations of ⁓ how many ⁓ images and videos ⁓ it equals to. So and yeah. And

Nathan Latka

23:45I see.

Jeff Lu

23:46Also on the pricing page just over there, it also tells you roughly how many videos and how many images you can create.

Nathan Latka

23:53So how many videos are you c are are your users altogether free and paid creating per day on a cool today?

Jeff Lu

23:58Yeah. ⁓ that's a lot. I think it's about ⁓ close to half a million, probably. More actually probably more than half a million. About half a million, I would say, ⁓ per day. Images, videos and all the assets ⁓ created per day on that code. So

Nathan Latka

24:19That's insane. I imagine that's just growing like crazy, huh?

Jeff Lu

24:23Right, right.

Nathan Latka

24:24Tell me more about how you've capitalised this as we wrap up. Is this bootstrapped or have you raised capital?

Jeff Lu

24:29⁓ we raised capital, but it's a small amount. So the company is largely ⁓ live ⁓ live by itself. So

Nathan Latka

24:38Okay. When did you raise capital and how little was it? What year?

Jeff Lu

24:42Yeah, so we we we raised in total about ten million dollars across the last four plus years. And ⁓ the last price round is about two years ago and we are ⁓ working a new price round as well.

Nathan Latka

25:03Interesting. Are you comfortable sharing what that what that valuation was in twenty twenty four, the last price round?

Jeff Lu

25:09Yeah, yeah, that's ⁓ that's around ⁓ hundred million dollars, two years ago. And ⁓ we haven't ⁓ haven't really raised the price wrong, but we take some safe notes at a much higher valuation and ⁓ we are thinking about a price wrong, you know.

Nathan Latka

25:14Okay. Okay. What what I mean, what's the market looking like today for valuations on companies growing as fast as yours? You're having conversations. I mean, do you think you can raise it a billion dollar valuation?

Jeff Lu

25:37so the market is constantly changing on the valuation side, ⁓ for the for the company. A billion may be a stretch, but I think probably like ⁓ half of it or something something something like that. But ⁓ a billion could be a little bit stretch on the current market, though yeah.

Nathan Latka

25:54If your old bosses at Google come back and say, Jeff, we want you back. Here's 500 million all cash up front. We want to acquire a cool. Do you take the deal?

Jeff Lu

26:05⁓ likely ⁓ I will. Yeah.

Nathan Latka

26:10I love that you're honest. Most founders go, No, Nathan, never. I want to build this forever. And you're like half a billion cash up front, I'll take it.

Jeff Lu

26:18So, well, I mean I I I I love the I love building business, I think definitely, but it's if there is a good opportunity, I I I I I my my big fan of the Elon Musk he keeps building new business, right? So

Nathan Latka

26:35Yep. Yep. All right. Well, Jeff, as we wrap up here, you look you're a technologist by training. ⁓ when you look at the technical side of pushing video creation forward, the next leap, what's the current bottleneck on the technical side?

Team Size and Sales Structure

Jeff Lu

26:48Yeah, yeah. The current bottleneck ⁓ I think ⁓ the videos are still pretty hard to create a long format and ⁓ realistic videos and so on. ⁓ and ⁓ how can we create one hour movie easily and how can we like make everyone able to create it? There's still some ⁓ some gaps over there. ⁓ and what we do is we ⁓ build the tools to make people very easily to create the long format videos that looks real and it looks good. And the the foundation model technology is growing pretty fast. So ⁓ think ⁓ pretty soon ⁓ people will be able to create ⁓ very good movie quality ⁓ videos by themselves.

Nathan Latka

27:37Can I ask you since you sit on top of other models and you're building your own, what are you spending per month in your cost to goods sold? Just out to you know, credits for other providers?

Jeff Lu

27:47Yeah. That's ⁓ that keep changing as well. So we used to be very low. when we depend on less third party models, but ⁓ when we use more third party models and so on, then cost of goods sold becomes pretty high. I think we are ⁓ cost of goods sold is over ⁓ million dollar a month these days. ⁓ and ⁓

Nathan Latka

28:12Yep. So how much how much of your time are you spending on inference models and tool calling to make sure that every sort of work request your customer asks for, you route to the cheapest credit on either your model or the foundation model or Hugging Face open source model?

Jeff Lu

28:26Right, right. So ⁓ we we actually let user choose and ⁓ user decide what model they want to use rather than we have a auto selection and ⁓ we help them to route. So currently it's user choose which model to use.

Nathan Latka

28:45I s would you ever change that? Don't users make bad decisions on that sometimes? Do you hold any patents around AI video creation? Very cool. Well, Jeff, before we wrap up, you've been very gracious with your time and very transparent. My audience has learned a ton. Is there anything else you wanted to cover that we didn't get a chance to cover? All right. Well hey, if people want to follow your story after this, where can they find you online? Guys, scientist from Google in 2022, he saw where Vidy was going and said, I'm quitting Google. I'm going all in on my side project, which at the time was only doing $100,000 per year. Fast forward to 2023 and a cool is launched. Turn your ideas into reality with next generation AI marketing, specifically around video. By 2023, they broke $3 million of revenue, land their first enterprise contract with Coca-Cola, which was bottom-up prosumer. To enterprise motion and by 2024 did 13 million of revenue and closed ⁓ finished closing 10 million total capital up to that point in time at a priced round of about 100 million. Now in 2025 broke 25 million last month. So we're recording in July, last month, which was June 2026, did $3 million of total sales, and users are now creating over half of million. Videos and images per day on his platform, 20 to 30,000 total paying customers. And he's doing he's doing all this with a team of 60 in a very capital efficient way. He's raised less total external capital than his total ARR, but rumor has it he's potentially looking at a larger price round today. We will see what happens. Jeff, thank you for taking us to the top. All right, guys. Cut Jeff, what'd you think? You have fun? Yeah, congratulations on what you've built. I mean, it must just be so fun every day looking at all the videos people are making on your platform and getting ideas. Yeah, that's great. Well, hey, I'm looking forward also to spending more time with you in Napa. I think people you're gonna enjoy meeting the other CEOs there. They're all doing thirty, forty, fifty, a hundred million bucks of revenue. It's gonna be a lot of fun. All right, Jeff. Talk soon. Bye bye.