How Ramsri Built Supermeme.ai into a $8500/Month AI Side Project Business

Introduction

AI side project business lessons from Ramsri on building Supermeme.ai and Question.ai to $8,500/month while keeping a full-time job. In this Q&A, he shares users, revenue, distribution, and his exact tech stack.

Founder Success Story QnA

Tell me a little bit about who you are, what you built, and what’s your story.

My name is Rams, and I build two apps that make me $8,500 per month. I’m currently building these two apps alongside my full-time job. Right now, I’m spending about 2 hours a week on these apps, and they mostly run on complete autopilot.

Tell me about some of the numbers behind these businesses.

Supermeme has done lifetime revenue of about 130k USD and Question is around 100k USD. Both of them combined are at $8,500, so approximately a little over 100k ARR. Coming to the users: 1.5 million plus users for Supermeme and about 150k plus users for Question. Paying customers: 400 for Supermeme and 180 to 200 for Question.

What is it like to build profitable projects while you have a full-time job?

The key during the initial phases of development of these two projects was to assign one to two hours of daily learning and development. SaaS has quite a lot of components—not just building the tech aspect of it—but there’s a whole suite of things: payments, authentication, databases, marketing, SEO. It took me a few years to gain proficiency in all these aspects. Consistency is the only thing that helped. Nowadays with tools like Deep Research and AI-assisted coding tools like Windsurf, Cursor, it has become a lot easier to build as well as maintain these apps alongside a full-time job. Nowadays I mostly treat my SaaS apps like they are in autopilot mode—just allocate Saturday afternoon, a couple of hours, to do updates or quick bug fixes.

Why haven’t you decided to go full-time?

I don’t want to take on VC funding or take on the responsibility of hiring and maintaining a team for the sake of growth. And several other aspects of life—for example, I have a kid, retired parents. I don’t want people around me to go through the same roller coaster ride of emotions and being always busy and unavailable lifestyle. I don’t want them to experience that.

Walk me through your background and how you got into building apps.

I was full on the MOOC revolution doing every technical course on Coursera, Udacity, and all these platforms during 2015 to 2018. When I started doing these, I started getting a lot of ideas in the AI space and I thought it’s about time for me to experiment. So 2018 to 2020, I just got my feet wet into entrepreneurship. I did not have any entrepreneurship background, so I was just testing the waters. I started writing all about my experiments on LinkedIn and Twitter. Eventually people started following me and asking questions, and waking up every day and writing online on these platforms became my point of comfort. During 2021 and 2022 I took all my learnings and launched a Udemy course and that made like $10,000 in revenue till now. In 2022–2023, I learned no-code tools and created the first version of Question.ai—no beautiful UI, just two text boxes left and right. Then I realized the bottlenecks and learned full stack with the latest tech—spent 6 months and fully ported Question to a modern tech stack. For Supermeme, since I knew the pain of growing a solo startup, I partnered with a full-stack developer—I’m the AI guy—and there’s another marketer. That worked out well. About a year, year and a half before, I took up a full-time job as well and I’m quite happy that it all worked out.

How did you find these ideas and what’s your framework for finding ideas worth building?

In short, to fit it in a framework: get manual tasks that people currently pay for, which could be significantly enhanced or substituted by existing or imminent generative AI. That’s where the key is. For example, with Question.ai back in 2020, we identified that large tech companies and publishing companies do a lot of quiz creation. We saw potential. AI wasn’t there yet to deliver automated quiz creation from text, but we could foresee that this could grow significantly in the next few years. You should always keep a tab on where generative AI is headed so that you can identify new business use cases that the new AI technology can unlock.

Tell me about growth and how you got your first users.

Build distribution before building a product. By distribution, I mean write about stuff that interests you or what you’re working on. For Question, before even building the product, I wrote a lot of blogs about quiz generation, created some YouTube videos. Every single day, I showed up on LinkedIn, writing about what I’m working on. Both Question.ai and Supermeme.ai were built in public. Sometimes I got 10 likes, sometimes 100, 400—the algorithm keeps changing, but consistency matters a lot. That’s how I even found co-founders; they saw my posts and we got together. We iterated on initial MVPs, then launched on Product Hunt. With the initial set of users from social media and the virality of an AI meme generator, we got to number one on Product Hunt on launch day for Supermeme.

Walk me through your tech stack and tools.

It’s Next.js for primary app development—a full-stack framework based on JavaScript. For backend, primarily Supabase. For landing pages we use Typedream to build no-code landing pages and do rapid SEO changes. There’s heavy development with AI-assisted coding tools like Windsurf and Cursor. This is our current tech stack.

What do the costs and margins look like?

We operate at a 70 to 75% profit margin. The biggest cost is mostly APIs—AI APIs from OpenAI, Anthropic. We have no employees aside from the founders. That helps us keep costs minimal.

What does a typical day in your life look like?

My day starts at 7:30 a.m.—wake up, get my kid dressed, drop him to school. By 9:30, I’m free. I walk on the treadmill and catch up on AI news. At 12:30 I pick up my kid, we have lunch till 2:00 p.m. Between 2 to 5:30 p.m. I do the bulk of my full-time job. Then dinner with family. If there are urgent tasks, I work 9:30 to 10:30/11:00 p.m. Otherwise, I watch TV or documentaries and go to bed by 11:30 or 12:00. I don’t pick up my SaaS stuff daily—I dedicate Saturday afternoon, usually 2 to 4 or 5 p.m., for feature updates or housekeeping tasks.

What advice would you give yourself when starting out to find that first app idea?

Think like a builder from day one. In today’s age, it’s possible to adapt that mindset and build something quick. Even if you want to do web coding, there are tools like Lovable and Bolt.new. If you want to use a chained workflow app, create automations with Make, n8n, or Zapier. When you have an issue or an idea, build a minimum viable prototype using these tools. Follow tech updates closely to identify new use cases. Aim for solving a problem by chaining and stitching together tools and products out there rather than spending months in R&D—that’s crucial.

Ramsri Business Stats

Here are the core business stats for Ramsri‘s AI side project business—covering users, revenue, paying customers, and margins across Supermeme.ai and Question.ai.

  • Monthly revenue: $8,500+ across two apps
  • ARR: a little over $100,000
  • Users: 1.5M+ (Supermeme), 150k+ (Question)
MetricValue
Monthly Revenue$8,500
ARR$100,000+
Supermeme Users1,500,000+
Question Users150,000+
Supermeme Paying Customers400
Question Paying Customers180–200
Profit Margin70–75%

Ramsri Method

Here’s how Ramsri systematically built his AI side project business from idea to distribution while working full-time.

  • Identify paid manual tasks that can be enhanced or replaced by generative AI.
  • Validate early by writing content and building distribution before the product.
  • Ship MVPs quickly using no-code and iterate to a full-stack modern stack.
  • Build in public daily on LinkedIn/Twitter to attract users and co-founders.
  • Leverage launch platforms like Product Hunt with early champions.

Ramsri Tools

Ramsri uses a lean toolset to build and maintain apps on autopilot—combining full-stack development, no-code landing pages, and AI coding assistants to keep speed high and costs low.

  • Next.js — full-stack app framework for rapid development.
  • Supabase — backend/auth/database services.
  • Typedream — no-code landing pages and fast SEO tweaks.
  • Windsurf & Cursor — AI-assisted coding for faster development.
  • OpenAI & Anthropic — AI APIs powering core features.

Key Notes

These are the standout takeaways from Ramsri‘s AI side project business journey that you can apply immediately.

  • Consistency in learning and shipping is the edge for solo builders.
  • Distribution first: write, post, and teach before you build.
  • Use existing tools and APIs to stitch together solutions fast.

Get Started in Just 5 Steps

To start your own AI side project business like Ramsri, follow these five practical steps and keep your scope small while you validate.

  • Pick a manual task people pay for and map how AI can replace or augment it.
  • Publish 5–10 posts and 2–3 articles to build distribution and collect feedback.
  • Ship an MVP using no-code; validate usage and refine the core workflow.
  • Port to Next.js and Supabase as you gain traction; keep landing pages in Typedream.
  • Launch publicly (e.g., Product Hunt) with your early audience as champions.

Conclusion

An AI side project business can thrive alongside a full-time job. By focusing on distribution, leveraging generative AI, and keeping operations lean with tools like Next.js, Supabase, and AI APIs, Ramsri grew Supermeme.ai and Question.ai to $8,500/month with 70–75% margins.