Namaste Tools

Top-quality video and image AI APIs, at the cheapest price

September 5, 2025 From tinkering with scrap GPUs to $5.8k MRR in 3 months (site still WIP 🙃)

Hey Indie Hackers 👋

I’m the maker behind namaste.tools, a set of AI tools for watermark removal, background removal, PDF cleanup, and image upscaling. What’s unusual is how I built it: on scrap GPUs powered by solar panels, stitched together with duct tape (metaphorically) and a lot of late-night hacking.

Three months in, I’ve reached $5.8k MRR… even though my site is still a very obvious work in progress. Here’s the story.

How it started

I didn’t come from an AI research background. My teacher was ChatGPT. I’d ask it questions, sometimes embarrassingly basic ones, and slowly learned how to fine-tune models, clean datasets, and get things production-ready.

Most people run AI workloads on shiny new GPUs in the cloud. That’s expensive. I didn’t have that luxury, so I went scrappy. I started collecting old GPUs, patched them into a cluster, and powered them with solar. Suddenly, I had a compute setup that was dirt cheap to run, eco-friendly, and surprisingly powerful.

The first customers

I launched quietly with a barely functional site. No big launch, no polished branding. Just a tool that solved a real problem: getting rid of watermarks and backgrounds cleanly and affordably.

Within 2 weeks, people started using it. A designer trying to clean up stock assets. A startup needing bulk background removal. Even a couple of agencies looking to upscale images for campaigns.

And then the questions came in:

“Why are your prices so low compared to other watermark removers?”
“Wait… solar-powered GPUs?”
“This works better than I expected.”

The numbers

3 months in: $5.8k MRR.
Site status: still half-broken in places.
Marketing budget: basically zero.

What I’ve learned so far

You don’t need a polished site to get customers. If it works and provides value, people will forgive the rough edges.
Scrappiness can be a superpower. Scrap GPUs + solar isn’t just cheap, it’s a differentiator.
MRR grows when you solve an everyday problem really well. Removing watermarks and backgrounds might sound “boring,” but it’s painfully useful to the right people.