In this insightful interview, Niklas Vesely sits down with Napoleon Biggs, co-founder of InstantStudio AI, to discuss how custom AI image generation software is transforming businesses across various industries. They dive into practical applications of AI technology, from automating photo shoots for fashion brands to creating personalized pet merchandise and enhancing recruitment processes with professional profile images. Napoleon shares the key considerations for companies looking to integrate AI tools into their workflows, highlighting the balance between innovation and practicality. Whether you’re a fast-moving entrepreneurial company or a business with growing demands, this interview breaks down how AI solutions can ease operational pressure, scale creativity, and unlock new opportunities.
Niklas Vesely:
Hey everyone, I’m Nick, helping out with the marketing here at InstantStudio AI. I’m joined today by Napoleon, one of our co-founders. Napoleon, can you give a quick introduction for those who may not be familiar with you?
Napoleon Biggs:
Hi, I’m Napoleon Biggs, sitting in our office here in Bangkok, though I frequently travel between Bangkok and Hong Kong. I’ve come on board to help build our client base and work with businesses that are looking to use custom AI image generation software to improve their operations. One of my main roles is to help communicate the benefits and capabilities of our AI image generation tools because, let’s face it, there’s a lot going on in the world of AI right now. Some of it is hyped up, but what we’re focused on is delivering real, practical solutions.
Niklas Vesely:
AI software can seem overwhelming, especially since it can do so much. Could you summarize what people can actually buy from InstantStudio AI, and what types of custom AI image generation software or image generation plugins are available?
Napoleon Biggs:
Absolutely, it’s really fascinating. I’ve been involved in software for over 20 years, witnessing the entire evolution of the internet. With new technology like AI, we often spend a lot of time explaining how businesses can use it effectively. What we’ve built at InstantStudio AI is simple yet powerful—custom AI image generation software that helps businesses create images on-brand. Our secure AI engine is hosted on private servers with our own GPUs, designed specifically for companies to produce highly precise, look-alike images, whether it’s for people, pets, or products. The goal is to help businesses generate images that promote their brand and products efficiently.
Niklas Vesely:
You mentioned that the images are very precise, but how does it actually work? I’ve seen many different AI tools on the internet. Some generate completely new images from scratch, like OpenAI’s tools, and others allow you to upload a photo and modify it. I’ve even heard of software that can learn a specific style from previous images. Which of these describes what your software does, or can it do all of them?
Napoleon Biggs:
We can actually do all of those things, but the focus isn’t on what the software can do—it’s about what people want to achieve with it. A year ago, when this technology first took off, everyone was going wild with it. It felt like being in a nightclub where you got to control everything—the music, the lighting, and what everyone was wearing. People were running with their wildest dreams, but now we’re coming back to a more practical approach.
What we’ve done is take existing AI image generation software from models like Stable Diffusion and Flux, along with new evolving technologies, and built a platform around it. For example, a business can come to us and say, “Here’s my fashion item—a leather jacket, a pair of jeans, or a sweater. I want this to be worn by a guy in his late 20s, standing on a street in Berlin in cold weather, for a casual photo shoot.” The software will produce that exact image.
The beauty of custom AI image generation software is that it allows for a lot of creativity. Some people use it in more open-ended ways, while others—especially in the fashion world—have specific use cases. Fashion brands might be launching a new collection and need to get product images for their e-commerce websites, social media, or even printed catalogs. And yes, this is not limited to digital formats. We’re seeing it used in print media and in-store displays as well.
Another imaginative use is within the pet industry. Since the COVID pandemic, the demand for pet-related products has exploded. Pet owners love seeing their pets in fun, creative images—maybe dressed as Superman or in other funny outfits. Our AI image generation software helps create these types of visuals without going overboard. What we’ve done is create a system that “tames” the AI’s capabilities. The problem with large language models (LLMs) is that they can hallucinate or generate unpredictable results, but we’ve built our software to ensure it remains safe for families and reliable for businesses. It’s tailored for companies that need to maintain brand consistency or pet companies that want fun, engaging images without distorting the subject.
Niklas Vesely:
You mentioned that it can be used by fashion designers to put specific clothes on specific people. Does that mean you can combine two photos? For example, if I had a photo of a model and another photo of a new jacket I designed, could you actually put them together?
Napoleon Biggs:
I wouldn’t say it’s for fashion designers specifically because that’s a very specialized skill set. A fashion designer typically starts with a drawing, and the materials, textures, and sensations are very important to their process. What we offer is further down the line—it’s more geared toward e-commerce managers and digital marketing teams. These are the people responsible for promoting and merchandising products through social media, creating high volumes of content for Instagram, sending out emails with new collections, and maintaining a consistent brand image.
It’s not a design tool per se, but rather a marketing and promotional tool. For example, in fast fashion, new collections are launched every few weeks, sometimes even daily with platforms like Temu. Keeping up with that content demand is extremely challenging. Marketers need to be able to generate images quickly and adapt them to different audiences. With our AI image generation software, you can create images that match the body type, ethnicity, or style preferences of your target audience, making it highly personalized. This allows for high-volume image generation without the need for traditional photoshoots.
Niklas Vesely:
So, if someone works for a large e-commerce platform and often uses the same models as part of their brand identity, and then they receive 100 new clothing designs from suppliers, could they use your software to place those clothes on their existing models without doing another photoshoot?
Napoleon Biggs:
Yes, exactly. One scenario we’re working with is companies that have a set collection of models that they identify with, though this is more common in fast fashion than in high-fashion. In the high-fashion world, you’d see recognizable catwalk models, but for many brands, particularly in the mid- to lower-tier markets, models aren’t typically recognizable. They often use non-glossy, everyday models for their e-commerce sites.
What’s interesting is that with AI-powered image generation tools, these companies can create digital versions of their models. This allows them to use the same models across multiple collections and campaigns without needing to constantly book new photoshoots. The digital versions of these models can be placed in various scenarios, wearing different clothes, with proper licensing in place. Models themselves can benefit from this by extending their availability across multiple brands, enabling them to work in more places at once—virtually. This is a game-changer for brands looking to keep up with the fast pace of product releases.
Niklas Vesely:
That sounds pretty incredible. Let’s say a modeling agency has a portfolio of models, and right now, a model can only attend one photoshoot per day. Each shoot is costly because you have to pay for the model’s time. Would it be possible for the agency to scale this up using AI image generation software? For example, could they send photos of the clothing to your platform, and your software could generate images of the models wearing the new clothes without needing an in-person photoshoot? That would be a game-changer for the modeling industry.
Napoleon Biggs:
Yes and no. You’re touching on something important about the traditional modeling industry. At the top, you have supermodels who are in high demand and whose physical presence is required at photoshoots. But below that, agencies have huge portfolios of up-and-coming models with different looks, and they are constantly being booked for shoots by designers. In high fashion, physical photoshoots are still the norm—models need to try on the clothes, and the designers make their selections based on that.
However, there is a whole new category, particularly in fast fashion, where brands don’t have the time or resources to hold constant photoshoots. These companies often need high volumes of images for social media, e-commerce websites, and even print materials. They are looking for custom AI image generation software that allows them to produce images quickly and affordably. They want everyday-looking people—someone who looks like an average person walking down the street or sitting in a café—so the demand for AI-generated models is growing rapidly.
With our AI image generation software, agencies can create virtual models who look like real people but don’t require the time and expense of traditional photoshoots. We’ve already been contacted by several modeling agencies interested in using AI-powered virtual models because they see the potential to scale. Some markets are starting to accept the idea of AI-generated models more quickly than others. And if you look at platforms like Instagram and Facebook, we spend a lot of time engaging with fashion and lifestyle content, some of which is already created by AI influencers who aren’t even real people. These virtual influencers have personalities, stories, and followings, and this is becoming an increasingly accepted direction in the fashion world.
Niklas Vesely:
When I visited your company, I spoke with your colleague Adeh DeSandies, and we briefly touched on a project involving a children’s book. Am I allowed to mention it? It was like a safari-style book, and I was amazed by the concept. Did I understand correctly that parents would be able to upload a photo of their child, and the kid would become the hero of the book? Like, the child’s image would replace the images on the pages, and they’d be the star?
Napoleon Biggs:
Yes, you’re spot on, though I wouldn’t necessarily call it just a book—it’s more than that. We live in a world where many theme parks—whether it’s a zoo or an amusement park with cartoon characters—have evolved into immersive entertainment experiences. Traditionally, these places engage visitors by offering merchandise and the opportunity to take photos with a life-sized version of their favorite character.
What’s exciting now is that some companies, especially in the wildlife and educational sectors, are leveraging AI-powered customization to take this further. They’re creating shows and interactive experiences where children can become part of the story. For example, a child could upload their photo, and our custom AI image generation software would stylize that photo to match the theme or character design.
We essentially teach our AI to understand the brand’s design language and character traits, and when someone uploads a photo of their child, the software transforms the child into that character. So the child isn’t just taking a photo with the character—they actually become the character. Whether it’s a safari explorer or an animated hero, the child ends up “wearing” the same outfit, and you get an animated version of them.
This can be used in many ways, from creating personalized souvenirs like postcards, fridge magnets, and t-shirts, to more interactive experiences at the theme park itself. It’s not just basic photo editing like placing someone on top of a mountain—this is far more sophisticated. The result is a 3D-style, animated version of the child as part of the story. Right now, this feature is mostly static, but we’re working on adding more animation in the future.
Niklas Vesely:
Let’s say I have a theme park, and we already have automatic photos taken whenever someone rides the roller coaster. Could this process be automated so that the photos are instantly transformed? For example, as soon as they ride the coaster, the photo could go through your system and be edited so the person becomes part of the story in the image, as you mentioned. Or would it still require manual input, like uploading the photo to the software and making changes step by step, similar to Photoshop? How does it work?
Napoleon Biggs:
That’s a great question. With all of these technologies, it’s not just about whether you can capture the image—it’s also about the customer’s willingness to share it and what’s done with it afterward. In places like Germany, for example, which is known for its strict privacy laws, there’s an extra layer of concern. So, when you think about automating the process, you also need to think about whether the customer is comfortable with having their image captured and used in a fun or interactive environment.
Typically, in theme parks today, you’ll see something like this: they’ll take your photo before or during the ride, and when you come down screaming, it snaps a shot of you. You then get the option to buy it or not. Privacy concerns aren’t always addressed upfront, but there’s usually some sort of opt-in.
What you’re describing—automating the process where someone’s image is instantly transformed into a part of the ride’s theme or story—is absolutely possible. If the camera is connected to a computer, and that computer is connected to our AI image generation software, it could instantly transform the image. You could have photos of guests looking like cowboys or any other character, right after they finish the ride. The tech is there, but you’d need to ensure the customer opts in to this process.
The other key issue is what happens to the image afterward. That’s why we’ve built our platform on private servers. Many of the companies we work with are very concerned about the security of their brand’s imagery and data. They also want to ensure the customer’s images are handled carefully. We’ve set up processes where the images are automatically deleted after a set period of time to address these concerns. As the technology continues to evolve, questions around privacy, data usage, and image rights will keep coming up. That’s why it’s important to make sure these issues are addressed proactively.
Niklas Vesely:
Whenever I try online software for image generation, it’s usually a long process. I upload a photo, then write a lengthy description, and after a few attempts, I get an image that’s close to what I wanted, but it takes a lot of trial and error. Is your process similar, or is it different?
Napoleon Biggs:
It’s extremely different. What you’re describing is how many people used AI image generation tools about a year ago, where you’d write a text prompt, upload an image, and hope for the best. What we’re working with now is the next generation of this technology, designed for businesses, not end consumers. The goal is to create AI-powered services or products that businesses can offer to their customers, with a much smoother and more automated process.
For example, in the fashion industry, you might want to promote a new product like a jacket or shoes. Instead of manually tweaking images, you can use our custom AI image generation software to automatically place the product in different environments and settings. Let’s say you’re selling pet products—you want to show how a dog food bowl looks with different types of dogs in various environments (hot, cold, indoors, outdoors). The AI software can automate this by using predefined templates or scenarios, whether that’s a beach, a mountain, or a cozy home.
What’s unique about our solution is that it’s built with a business focus in mind. You don’t need to manually enter text prompts for every change. The system learns as your team uses it. If five or ten people from your company are working on it, they can build on each other’s knowledge. It’s like a form of AI knowledge management, where the system remembers preferences and settings from past uses, making it easy to generate consistent results over time.
For example, in industries like advertising or programmatic ad targeting, companies want personalized, localized images for different markets. A furniture company might need to show an older couple sitting on a sofa for one market and a younger couple for another. Our software allows for rapid high-volume image generation, removing the need for repetitive photoshoots and manual editing.
Niklas Vesely:
That sounds like it could save a lot of time, especially if you’re dealing with different product variations, colors, or age groups. Having to set up a photoshoot for each scenario would be impossible.
Napoleon Biggs:
Exactly. We’re compressing the time it takes to generate personalized images, and that’s a huge benefit. The ability to create personalized, high-volume content quickly is crucial in many industries, from fashion to pet products, to streaming video services. For instance, one of our clients produces a lot of thumbnails for their films, and they need to localize those images for different markets. Normally, this would be a repetitive, manual process. With our software, it’s automated and far more efficient.
There’s also a creative side to this. We’ve worked with designers—like a stationery company—that want to generate variants of their designs. They upload their artwork, and our AI-powered design tool allows others to create multiple versions of it, whether for birthday cards or t-shirts. This way, the creativity of human designers is enhanced by the AI’s ability to scale their work across different formats.
This version incorporates more varied keywords like AI-driven product personalization, automated image generation, and high-volume image creation to improve search engine visibility for companies looking for such solutions. Let me know if any further changes are needed!
Niklas Vesely:
There’s one story I heard that paints AI in a somewhat negative light, but it’s fascinating. A designer was fired after 10 years because his company fed all his artwork into an AI. The AI was then able to continue creating advertising content in his style, making his role redundant. Would your software be able to do something like that?
Napoleon Biggs:
Actually, yes. Any AI image generation software—particularly those built on large language models (LLMs)—has the capacity to learn from reference material. The AI learns patterns, styles, and techniques from the data it’s trained on, and then it can replicate or build on that. That’s the core functionality of this kind of technology. It’s also what makes it potentially concerning, as in the case you mentioned.
The real question for businesses is whether they want to use AI in that way. There’s an evolving conversation, much like when the internet first started, about how creators should be rewarded when their work is used by AI. On a larger scale, you’re seeing lawsuits against the foundational LLM models from companies like the New York Times and Reddit because they believe their intellectual property (IP) was used without consent to develop these technologies.
We’re also starting to see more regulations, especially in places like the EU, that aim to address these concerns. But it’s not just about restricting AI use; it’s about figuring out how to fairly recognize and reward the creators whose work is being leveraged by these models. This is critical when AI begins replicating an artist’s or designer’s unique style. Who owns that style now? Is it the company or the creator? What rights does each party have moving forward?
I’ve signed employment contracts myself where, in certain industries like advertising or marketing, every idea or design you create while employed belongs to the company. That’s standard practice. But as AI becomes more integrated into design workflows, there’s a need to re-evaluate how we reward creators, particularly when their look, feel, or style is being replicated by AI.
There are technologies like watermarking or blockchain that can help track usage and ensure the creator is compensated, but it’s still a developing area. When you use AI-powered software to create images, everything is trackable—there’s always a record of what’s been generated. The question remains: what happens to that image afterward? That’s where the murkiness comes in, and it’s something that the industry is still figuring out.
Niklas Vesely:
I remember visiting your company, and there was a use case we discussed that seemed quite simple, but it really helps explain how the software works. For example, imagine LinkedIn allowing users to upload a selfie, and the system could automatically generate a professional-looking photo where they’re wearing a suit, with a well-groomed hairstyle and background. Many people don’t have time to go for a professional photoshoot when signing up for something like LinkedIn. Could your AI-powered image generation software handle that? Could LinkedIn deploy it on that scale so users could just upload their selfies, and the system would handle the rest?
Napoleon Biggs:
Yes, LinkedIn is a great example. But the cases we’ve seen are more commonly with recruitment firms. In places like Japan, there’s a certain expectation of how you present yourself on a resume, and professional photos play a big part. You’re right—most people, especially parents, don’t have professional photos of themselves. They’re often weird angles, group photos, or casual snapshots.
What’s interesting about using AI image generation in this context is that it allows people to upload casual selfies and quickly transform them into polished, professional images. The system could, for example, put the user in a suit, adjust the lighting, and clean up the background.
In fact, we’ve noticed that a lot of AI tools tend to beautify images a bit too much, making people look like movie stars. But what we’ve focused on is something more realistic—people want to look like themselves, just in a professional setting. Our software can make subtle adjustments while keeping the individual’s unique features intact, allowing them to look their best without looking overly edited.
Niklas Vesely:
Amazing. So, what would it look like from LinkedIn’s perspective if they wanted to integrate this AI image generation software into their platform?
Napoleon Biggs:
In an ideal scenario, LinkedIn—or any company—wouldn’t even notice we’re there. It would be seamless, much like when you log into a website using Google Connect or pay for something through Shopify. Users don’t see the layers of technology behind the scenes.
If LinkedIn integrated our system, it could be as simple as asking the user for permission to access their existing photos. The AI could then generate a series of professional-looking images, offering choices like “Do you want to be in a suit?” or “Do you prefer a sweater?” The user could make their selection, and they’d have a polished photo ready for their profile.
From a company’s perspective, it’s all about seamless user experience. We work with companies that really understand their customers—whether it’s a mobile app, e-commerce platform like Shopify, or even a gaming service. Our role is to plug into their existing processes, ensuring the AI functionality is smooth and easy to use. For instance, in the pet product industry, users might upload a picture of their pet, and within days, receive a personalized phone cover with their pet wearing a cute outfit. The AI makes the user journey effortless, and we handle the backend integration.
Niklas Vesely:
May I understand what the process looks like for the user? And I’m also curious about what it looks like for the company that hires you to do this. Let’s say I run a job board or a job website similar to LinkedIn, and I want to add this AI image generation feature to gain a competitive edge. What does the process look like from my side? How complicated is it to get started—what would I need to do first, and what happens next?
Napoleon Biggs:
Ah, you mean the onboarding process? First, you just need to reach out and say hello. This is a B2B service, so the process is very consultative, at least for now. Because AI image generation is still evolving and every business has unique needs, it’s important to start with a one-on-one conversation. We work to understand what your business is trying to achieve and what kind of solution you need.
For example, you’d tell us what features you’re looking to add—whether that’s creating professional images from selfies, automating product photo generation, or any other specific application. Our job is to listen and identify the AI tools and technology that fit your goals. It’s a very personalized process at this stage.
In the future, I expect it will become more self-service as the technology matures, but right now, it’s still new and very custom. Much of the process is about educating businesses on how this AI software can meet their needs. We often spend around 70% of our time explaining what the technology can do, as many businesses are still figuring out how AI fits into their operations.
Once we’ve identified the solution, we’ll adjust the AI system to fit your specific requirements. After that, we build it, test it, and then pass it over to you. From there, it becomes a self-service workflow. Your team can log in, use the system, and generate the images, ads, product photos, or whatever else you need, based on the specific AI solution we’ve created for you.
So, it starts with a consultative approach, but once it’s set up, the system is easy for your team to use in-house to create outputs whenever they need. Whether it’s professional photos for job seekers, pet merchandise images, or AI-generated ad creatives, the system is built to empower your team to generate high-quality content on demand.
Niklas Vesely:
What does a suitable customer look like for you? Does it have to be a company with 10,000 employees, or can it be smaller? How does a customer know if they’re the right fit for you?
Napoleon Biggs:
That’s a great question, and I wish there were a simple answer! We’ve worked with companies of all shapes and sizes—from multinational brands to regional marketing firms and technology companies. At the moment, the most open to adopting AI image generation software seem to fall into two categories.
First, we see a lot of interest from entrepreneurial companies. These are the fast-moving businesses that already have a product or service and see the potential to use AI technology to enhance what they offer. They’re agile, not afraid of new technology, and they move quickly to stay ahead of the curve.
The second category includes smaller teams, often in regions like Asia, where businesses are under pressure to handle large volumes of work. These teams are responsible for managing operations across multiple markets and cultures, with different languages and expectations. They’re stretched thin and are looking for solutions that can ease some of that pressure. For these teams, integrating AI-powered solutions helps them scale their efforts without needing to hire additional staff or resources.
You’ll often find these types of businesses in cities like Singapore, Hong Kong, New York, London, and Paris—central hubs where teams are managing regional or global markets and just can’t keep up with the demand.
So, whether it’s a fast-moving, entrepreneurial company that sees AI as a way to create a new service, or a team that’s simply bursting at the seams and needs help managing their workload, we can offer solutions tailored to their needs.
Niklas Vesely:
Okay, amazing! Now I can see how versatile your offering is, and I’m sure it’s clearer for the audience too. Thank you so much for taking the time to explain all this.
Napoleon Biggs:
Thank you! You know, the hardest part is narrowing down what’s possible for each company. This AI technology can go in so many directions. The challenge is in cutting through the noise—there’s so much hype around AI that people think it’s going to replace everything, but in reality, it still requires human input and guidance. It’s about making it relevant to businesses in a practical, manageable way.
Niklas Vesely:
Before we started talking, it felt like such a big picture—kind of overwhelming and not very practical. But now, it’s much clearer how this can be used in a real-world context.
Napoleon Biggs:
Exactly. I’m narrowing it down on purpose because when you work with companies, you can’t just walk in and say, “I’m going to solve all your problems with AI.” You’ve got to address specific needs, focus on particular pain points, and offer solutions that are actionable. Otherwise, people get overwhelmed, and you end up solving nothing.
Niklas Vesely:
True, true. Napoleon, thank you so much for your time and insights!
Napoleon Biggs:
Thank you, Niklas! Talk to you soon!
Leave a Reply