I'm cutting my monthly costs by more than 80% by self-hosting everything

I started indiehacking a year ago & was using mostly free services for my projects:

- Vercel

- Supabase

- PostHog analytics

- AWS EC2 first year free for micro instance

But over the last year, my projects have finally started getting some more traffic + my free credits on AWS ran out. I am looking to scale even more so I know costs can get out of hand pretty quickly (+ I have seen some scary stories here recently).

Monthly, I would have, on average:

- $50 for Supabase (5+ projects)

- $45 for my AWS EC2 instances

+ variable costs for renting GPU for my AI projects

+ I also just surpassed the limits of the free plan on Vercel in terms of bandwidth, Image Optimization & Speed Insight points

The only project that's (finally) making some money so far is ImagenMIA, and with only 60% margins or so due to high model training costs (+ still early phase).

On the one hand I am super happy that ImagenMIA is starting to generate some money on its own (all organic, SEO-based traffic) but on the other hand, infra costs are starting to climb up and are only going to get worse as it scales.

Luckily I saw a post from @dobroslav_dev mentioning Hetzner and Coolify. So I started digging into it and have migrated the first couple of projects to test everything out.

Here is my setup:

- One small VPS (€3.98) for Coolify

- 2vCPU

- 4GB RAM

- 40GB SSD

- 20 TB Egress traffic

A bigger one (€14.51) to host all my projects:

- 8 vCPU

- 16 GB RAM

- 160 GB SSD

- 20 TB Egress traffic

I will be self hosting everything:

- Supabase

- Front-end

- NodeJS Express API Endpoints

- N8N

- PostHog

I was a bit afraid but tbh, Coolify makes things so much easier (OS alternative to Vercel/Netlify) + now I no longer have to worry about extreme bandwidth costs in case of traffic spikes!

Looking to migrate my bigger projects over there hopefully the next couple of days!

Finally, I managed to significantly improve image generation quality. 🎉

After countless iterations and days/weeks I finally managed to significantly improve my models. Models were good for some but quality was not nearly as good as I would want. Did countless iterations of Dreambooth training settings and managed to increase resemblance significantly.

Still some finetuning left to do but will be able to start providing significantly better images & pivot towards photorealistic CV/curriculum pictures soon!

On top of that, the first customers that have used the new model were very happy with their results 😄

✅ My gameplan next:

1. Image optimization & compression on the backend + filtering on frontend

   • Currently still stored as PNG and big size - slow to load + consumes too much bandwidth

   • Duplicate images check

   • Corrupted images check

   • Filter and check for bad images on client-side

   • Add big modal explaining bad input = bad result. Still too many ppl send very bad images.

2. Write and add 20-40 high quality prompts on top of the current ones to generate a variety of CV images (80 to 160 images)

   • Add styles to landing page & highlight more

   • Add styles to Uploading section + add select all + add possibility to select type (profile pics vs. avatars/anime/ ...)

3. Start email marketing

   • 7k emails from current users who have never been contacted. Time to change that

   • Goal would be to start a bi-weekly newsletter with info about AI models + improvements on ImagenMIA + discounts + refer to InterioresIA too

   • Need to have at least mails for the first 4 editions before I start the campaign - to allow for a good buffer

4. Start running ads

   • Purchased a course and will start consuming it & sharing notes

5. Programmatic SEO per image & background (similar to what I did for InterioresIA.com)

   • Step up the SEO game even more

   • Improve copy

Spent the last two days trying to improve the models 😅

And man, this shit is hard. Been doing a lot of iterations on training settings and playing around with parameters. I have managed to make some slight improvements, but far from being where I want my models to be.

Also spent a lot of time prompting to generate new and good prompts for professional photoshoots/CV pictures.

That's much harder than avatar profile pictures. With avatar profile pictures you have some room for error since it's an avatar/cartoon/anime version of the user. Here you don't!

Will keep at it and let's see where we get! This is important to be able to pivot into more CV type pictures, which is what visitors of the site obviously want!

Will also be good for me as I will be able to start charging more. Currently at 5-6$ per sale does not leave me with a good margin.

Other to-do's:

• Add image filtering on client side

• Failed models error handling

• Start mailing

Have been neglecting IndieLogs a little bit as I'm focused on ImagenMIA & InterioresIA

The next step for IndieLogs is to basically get more visibility, exposure and traffic. Will start posting on different repositories and also schedule a ProductHunt launch in the next month.

It's all about marketing now for IndieLogs. Have noticed my posts about it on X are barely getting any views compared to early during launch. I guess the 'wow' effect of it is gone or I've just become really bad at X lol 😜

In any case will make sure to find a way to get a good amount of traffic for IndieLogs as well. And at the end of the day, it's something I really like using to document my own journey - so it's always a win.

Almost ready for launch! 🚀

The actual core feature has been ready since day one. I spent the rest of the time working on the landing page, dynamic pages and making sure I was building everything targeting specific SEO keywords.

✅ Done so far:

• Over 350 SEO-optimized dynamic pages targeting different styles and rooms

• 19 room types supported

• 15+ styles

• Supabase

• Analytics

• Blog

• LemonSqueezy test

• n8n workflow integrations

• Landing page 90% ready

• Payment & credit system up and functional

• Backend & image generation up and functional

• Backlinks built using other projects

👇🏼 Still to do:

• Waiting for LemonSqueezy store activation

• Downloads +zoom on generation page

• Sitemap and indexing

• Refine and finetune styles

• Logo

• Generate blogs for programmatic SEO

• Upscale some before images

• Scroll down to generated image on mobile

• Convert images to JPG + compress them to improve SEO

• GSC and Ahrefs integration

Project progress going well! 🔥

Will be focussing on building a single feature - the most important one. After that, will let SEO kick in and not work on it for a little while so I can move on to other projects

✅The core MVP for this project would contain:

• SEO optimized landing page

• Programmatic SEO targeting long-tail keywords (articles + image displays) - aiming for around 1000 pages.

• The backend image generation engine (almost ready)

• Payments
  

Will avoid building too much before the idea has been validated. Only after seeing a good traffic spike + initial sales, will I further optimize the project.

🚫 That means:

• No image storage

• No gallery/dashboard

• No super-advanced image-processing engine

• No upscalers

• ...
  
  
  

Trying to avoid previous mistakes and keeping scope limited!

The changes I deployed last week made conversion worse, not better.

In a rush before my vacation last week, I redesigned the entire website and I deployed the updates with the end goal of boosting conversion rates.

But the opposite occurred. Not a single sale during the entire week. I wasn't too bothered about it as I thought it was just a random drop/seasonality effects or something similar. Until my sister came to visit and I asked her to test out my website.

That's when I realized there were 2 issues:

• It wasn't rendering properly on older browsers (OKLCH not supported)

• Google Auth was no longer working for new users This resulted in lower activity on the website, less sign-ups, and less sales.

I did the following to fix this:

• Re-implemented Google Auth integration with Supabase

• Added fallback CSS code to my TailwindCSS file to handle fallback scenarios for DaisyUI on older browsers
  
  

Result: woke up to 3 sales and currently have 5 sales in the last 24 hours! 🔥

Planning a pivot towards professional headshots 📸

Profile pictures are cool, but nobody pays for them. Also quite clear from the data that this is what people want:

• Every customer has purchased the realistic package. Other packages have 0 sales.

• All Google traffic is targetting keywords related to professional headshots.
  
  

I'm getting some good traffic from Google, and am pretty sure I can monetize it a lot better than I currently am. I have known that I needed to pivot for a while but I was too lazy/distracted with other projects to actually do it. Now it's time to really double down on this project as it's generating the most revenue/getting most traffic.

✅ Action plan:

• Short term:

  • New styles - have a variety of corporate headshot styles for people to use

  • Update landing page to reflect pivot

  • Update pricing to reflect change

  • Get rid of studio idea - instead focus on simplicity

• Long term

  • Improve ressemblance for ControlNet

  • Add poses, styles, backgrounds

  • Improve processing backend

Will work on the pivot and implementing the changes first before I start doing email marketing. I expect conversion to be higher then.

For tomorrow, these are the actions I'd like to take:

• Start experimenting with the new styles - have at least 10 combinations

• New landing page:

  • Pricing

  • Testimonials