Building one of the first AI-headshot products and hitting $900k/mo in 3 years

Wesley Tian built one of the first AI headshot products — Aragon. Three years later, he's bringing in over $10M/yr.

Here's Wesley on how he did it. 👇

Going all in

I wanted to build a tech company from a young age. My dad works in tech and I love challenges. Plus, it's the only way to really make money.

I've been really into AI for years. I applied to OpenAI back in 2017 and also took part in competitive machine learning competitions. Since then, I've worked at two early-stage YC-backed startups.

When I turned 25, I decided I wanted to live a life I didn't regret, so I started exploring ideas part time.

Around this time, I heard about Stable Diffusion, an open-source text-to-image AI model, and a research paper called Dreambooth, an algorithm used to teach an AI model a certain concept like somebody's face. After playing around with it, my cofounder and I realized it was going to be an inflection point in technology because of how much more advanced it was than the previous generation of text-to-image AI models like GANs (generative adversarial networks). It was a great business opportunity.

With $15k-$20k in savings, I started launching apps full time to see what people wanted. I soon had a site with 12 "apps" with waitlists. People liked the apps where you could change your appearance.

That led me to launch an AI avatar product with different styles. And people liked the professional headshots style, so I doubled down on that.

Building a $10M/yr AI product

My product is Aragon, a suite of AI photo editing and generation products. Our most popular product, an AI headshot generator, allows you to generate pro headshots in minutes, while choosing outfits, backgrounds, etc. We made $4.5M in 2024, and we're currently at $900k/mo.

Essentially, we use AI to generate hyperrealistic photos of people, eliminating the need for in-person photoshoots. Professional headshot photo sessions cost like $250 in the US and it’s so much hassle — booking an appointment, getting there, putting on a suit/makeup, etc. And you only get one outfit and background.

We built the prototype in a day using Gradio. Our stack was React, NextJS, TypeScript, Postgres, Heroku, AWS, GraphQL, and it hasn't changed much.

We charge between $35 and $75 for headshots, depending on the package and number of headshots you want. We also sell to businesses, from SMBs to enterprises. SMBs purchase credits and we create custom contracts for enterprises, depending on their requirements. We also have a subscription-based AI photo editor that costs $15/mo ($129/yr) and a subscription-based AI photo generator that costs $29/mo.

Our margins are quite good; our primary costs are salaries, GPU costs, and marketing costs such as running ads.

We have over 1.3 million users, including Fortune 500 companies, and our products have generated over 30 million photos for people.

Hiring is hard

Hiring talent is pretty much the hardest thing I've done. We've fired so many people and wasted so much money working with agencies.

Generally, my advice is to avoid working with agencies and part-time people. And learn a role before you hire for it. It'll allow you to know what to look for and determine if they know what they're talking about.

Just like with products, hiring processes require constant iteration.

Increase your "luck surface area"

We grew through guerilla marketing, building in public, and our affiliate program.

Build a good product

It's very important to have a good landing page. But the most important thing is building a good product.

Launch frequently

I've launched multiple times. Pro headshots launch was in February 2023 when there was only one other competitor, TryitonAI, which shut down.

We launched on X, LinkedIn, and reddit. That launch made me $3k, but most of my launches early on only made a few hundred.

Try guerilla marketing

Guerilla marketing helped in the early days. I posted on reddit.com/r/tinder saying that I got a date because I used Aragon to generate better pics of myself. I posted on Teamblind saying that I got a job because I used Aragon to create headshots of myself. I

Leverage social media

I built an audience on X and LinkedIn that gave my product a lot more exposure. It increases serendipity. So I post everywhere. I have no shame.

Don't just list what you did. Also talk about how you did it and why it worked or failed.

Just be authentic. If you become successful, you'll have lovers/haters anyway. Just own it. But don't intentionally a dick.

It's all about getting eyeballs on your product while increasing your luck surface area.

Get affiliates

When we started an affiliate marketing program using Rewardful, people started linking to us organically.

That in turn helped us rank higher on Google, which brought in a lot of traffic.

Focus on your CRO

Conversion Rate Optimization is always important. Spend time on it and A/B test.

What didn't work

Cold emails did not work for me at all. Meta ads didn't work either. And PR was a waste of time.

Personal advantages

It's really important to have a routine and discipline. Read Atomic Habits.

It also helped that I’m pretty competitive. And I'm a generalist, so I enjoy being challenged and learning new things.

And then, of course, there were the forces outside of my control. Luck, timing, and market trends play a big role.

For example, a popular mobile app launched an AI headshot feature and it went viral on TikTok. Because we were ranked highly on Google, we got a lot of traffic and purchases from them.

To be fair, though, timing is under your control. To a degree, at least. Look at new technologies coming out and think about how they can be applied. How can they solve existing problems faster, cheaper, etc. than before?

Parting advice

Here's my advice:

• You have to build something that people want. And you need them to want to pay for it.

• Launch early. Most founders don't launch early enough. Our first landing page was terrible, but if we hadn't launched when we did, things could be different

• Move fast. When new tech comes out, test it, adopt it, and ship it quickly. Don't slow down as you scale.

• Keep your expenses low and your margins high.

• Don't get blinded by numbers and vanity metrics.

• Be on X and LinkedIn. Stay on top of all the new things coming out if you're early in your journey. You never know what might inspire you.

• Your network is important. It'll help you in ways you can't foresee. If you can't physically be in San Francisco, follow all the people there, or work for startups in the startup/VC world.

• I wasn't good about exercising and getting enough sleep all the time, but of course, I recommend it. Exercise 5+ days/wk, avoid junk food, and go to sleep and wake up at the same time every day.

• Focus on your personal life. Having a stable life outside of work helps you to weather problems at work. I didn't do this, and it's the reason I burned out twice.

• Schedule time every day to not work. And take vacations. Time off is essential. It's the only way to truly recharge.

• And find people who believe in you, you only need a few. For me, it's my parents, ex, investor, and friends.

Quality resources

X is incredibly helpful. If you're not on X, get on there. That's where we initially learned about Stable Diffusion and Dreambooth, the core technologies that led me to work on Aragon full-time. It's arguably the reason for our success. And I also learned a ton from famous indie hackers such as Pieter Levels.

I've also learned a lot from fellow founders in San Francisco.

And Atomic Habits is a must read.

What's next?

In terms of company goals, we're looking to grow revenue significantly by spending more time on our other products, such as our AI photo editor and AI photo generator.

We want to improve conversion, retention, and UX. And we're looking to hire a few more people, build more products, and scale up certain growth channels.

The biggest bottleneck for achieving my company goals will probably be finding and hiring the right people.

Personally, I'm looking to continue to travel and meet people and also improve my VO2 max (maximal oxygen intake) and get into the best shape of my life by doing more HIIT and cardio.

And I'll probably start another company in the future, probably in deep tech.

You can follow along on X, LinkedIn, and my website. And check out Aragon!