Building an OF-Style Creator Platform, Tools & Automation – Starting From Zero

Hi everyone,

I thought I would start documenting my journey building AlbumSelection, an OF-style creator platform that I've developed from scratch over the past year as a solo developer.

The idea came from observing how difficult it is for many creators to consistently grow their audience, gain subscribers and turn content creation into a sustainable income. I wanted to create a platform that incorporates some of what agencies do, along with tools and automation to help creators grow, monitor performance and reduce manual work.

Rather than building just another subscription platform, I wanted to build an ecosystem around creators.

So far, this has included:

• Building the platform itself from scratch in Python (a big task!)
• Designing creator profiles, subscriptions and PPV features.
• Creating onboarding and qualification systems.
• Setting up automations to reduce manual work.
• Experimenting with traffic acquisition and creator recruitment.

I'm documenting the journey here because I enjoy building things, experimenting, sharing results (good and bad), and hopefully connecting with people who are interested in creator businesses, AI, automation, traffic generation or growth systems.

The project is still very much a work in progress, and I expect there will be plenty of mistakes, pivots and lessons learned along the way.

I'll try to post updates regularly and share what works, what doesn't, and where the project goes from here.

My tech stack

I thought I'd share what I'm using to build AlbumSelection.

Current stack:

• Python – Main programming language. I've been using Python for years and wanted to stay productive rather than chase the latest frameworks.

• py4web – Chosen because I like its simplicity, batteries-included approach and rapid development capabilities. It has allowed me to move quickly as a solo developer. I use PyCharm as my development environment.

• HTMX – One of my favourite discoveries. It lets me create dynamic interfaces without writing a huge amount of JavaScript.

• Bulma – Lightweight CSS framework that keeps things clean and easy to maintain.

• AWS – Hosting the platform and related services.

• Amazon SES – Used for transactional emails, onboarding emails and lead follow-ups.

One thing I've learned while building this project is that, as a solo developer, shipping features quickly is often more important than having the "perfect" stack.

The stack may evolve over time, but so far it has allowed me to go from an idea to a working platform relatively quickly.

I'm always interested in hearing what other solo builders are using for their projects.

Developing code – probably my biggest challenge.

While developing the basics for the platform, I knew I wanted to include a live chat feature. I soon realized that this was beyond my capabilities at the time.

I turned to Upwork and found an excellent professor from Ukraine who was doing Python side projects. I hired him to develop the chat capability.

My mistake was that he developed it as a standalone py4web application, and I then had to spend time integrating it into my own application. Looking back, what I should have done was provide him with a copy of my existing project and ask him to integrate the chat directly into it.

That was a big lesson and probably cost me a couple of weeks of development time.

What worked out really well, though, was having live sessions with him. He helped me set everything up correctly, improve my workflow, and understand how a more experienced developer approaches projects.

By the end of it, I felt like I was finally working like a real developer rather than just hacking things together and hoping for the best.

Lesson learned: when hiring someone to build a feature, always give them the actual project they'll be integrating into, not just a specification for a standalone module.

Developing with AI

One thing that significantly accelerated development was using ChatGPT (which I affectionately call superNova ).

Rather than replacing development skills, I found it acted more like a patient senior developer sitting next to me.

It helped me:

• Understand py4web concepts
• Build controllers and forms
• Debug issues
• Improve SQL queries
• Refactor code
• Brainstorm creator acquisition funnels
• Generate translations
• Draft emails and onboarding sequences
• Review ideas before implementing them

The biggest benefit wasn't necessarily writing code. It was reducing context switching and helping me stay productive as a solo founder.

I still make mistakes and sometimes spend hours debugging things, but AI has probably compressed months of learning into weeks.

All of a sudden, I felt like I had gone from being a beginner-average coder to having a senior developer with 20 years of web development experience sitting beside me, available 24/7. It genuinely felt like having a superpower.

Of course, AI is not perfect. It can make mistakes, go down rabbit holes, and occasionally suggest things that simply don't work. But when used as a development partner rather than a replacement for thinking, it has been one of the most powerful tools I've ever used.

Challenge: Finding Creators

After spending months building the platform, I eventually faced a bigger challenge: finding creators.

To be honest, I initially recruited manually because I had absolutely no idea what I was doing .

I managed to onboard my first 12 creators relatively quickly. Some came through friends, while others I found on sugar-daddy websites and reached out to directly to see if they would be interested in joining.

I learned a lot from those first conversations.

Some creators wanted me to buy the rights to their existing content outright. Others expected a fixed salary simply for joining the platform. A few signed up enthusiastically but then disappeared once they realized they actually had to create content consistently to make money.

That was probably one of my biggest early lessons.

I realized I wasn't looking for creators who wanted to be employees. I needed creators who were willing to invest time and effort into building their own business, with me providing the platform, tools, automation and support to help them grow.

Looking back, onboarding those first 12 creators manually was probably worth it. Even though most didn't stick around, I learned more about creator psychology and expectations than I ever could have from reading articles or watching YouTube videos.

Sometimes the fastest way to learn is simply to jump in, make mistakes and adjust as you go.

Sounds good, will be interesting to watch the progress.

Building the Recruitment Funnel

One thing I quickly learned was that building the platform was probably easier than recruiting creators.

I went through three iterations of the recruitment funnel before arriving at the current version.

The first version was very basic and mostly involved manual outreach. I was essentially figuring things out as I went along.

The second version introduced more structure, qualification and onboarding, but I still wasn't attracting enough interest from the right creators.

The third iteration is the current one. The biggest change wasn't technical - it was realizing that my initial offer simply wasn't compelling enough.

After talking to creators and seeing where people dropped off, I decided to make the offer more generous and focus more on the value creators would receive upfront.

That significantly improved interest and applications.

The funnel is still evolving, but one thing I've learned is that you shouldn't become emotionally attached to your first offer. If people aren't responding, it's usually a sign that you need to test, adjust and improve until you find something that resonates.

I'm sure this won't be the final version either, but for now, it feels like I'm moving in the right direction.