Cal Counter by Metabola

Take a photo of your food and have your calories tracked by AI. There are numerous copies, notably Cal AI which sold for $50 million dollars. A small fraction of the market share is very valuable

Comment

Inspiration

I was experimenting with new tech stacks while starting a new diet. It took me over 30 seconds to just start logging my weight or a food item from competitor apps, so to save time I decided to just write my own. After a couple of days I had something that worked for me, but then I decided to see if I could publish it. After over 4 months of nights and weekends I was able to get it into a production ready state.

What it does

It's a seamless way to track your diet and weight. It uses AI to automate the calorie counting, or to at least give you a starting point which you can edit or tweak manually.

How we built it

I built it by getting some basic user flows to work end to end across RN, FastApi, and Supabase then I went on and added the rest of the features it has today.

I used AI to do most of the coding, but I would read every line of code. I started this with Gemini Pro 2, then to Sonnet 4.6, Fable when it was available at a discount and now Opus 5. It was very interesting using different models because I felt Gemini 2 wasn't able to cut it, it needed significant manual intervention. Fable was a pleasure to work with, it was able to resolve some very difficult bugs. Things like one component failing to unload, visually it's not there, but it would still eat inputs in certain parts of the screen.

Challenges we ran into

The biggest challenge is/was marketing. I enjoy building so while there were major issues along the way, like getting biometrics to work flawlessly and seamlessly, then just to remove biometrics because I realized all major competitors aren't locking the user's account from inactivity and biometrics would just be an extra step every time.

Accomplishments that we're proud of

All of the great feedback I've received on the app so far

What we learned

Too many things to list, but I've gotten very familiar with the apple app ecosystem. I learned a ton about marketing, sales and general business skills.

For the technical side, I'm coming from enterprise .NET, so everything in the stack for this app was new to me starting with this project.

What's next for Cal Counter by Metabola

The next steps are to hit a holding pattern after App Store review. Verify everything is working as expected once live. I'll then focus the dev work on fully automating the build + release process. Get a dedicated dev/staging environment setup and just start cleaning up other tech debt.

From there it's marketing and advertising I really want to fully understand my ideal customer profile and have a significant amount of practice in organic content for advertising. For the fastest scaling I would go with paid UGC, but this is 100% a solo project while I work full-time and I'm working with a very limited budget, so I need to get enough organic traction to calculate my CAC before I step into paid marketing.

I really hope to hit a breakeven relatively quickly then I can grow slow and steadily.

Built With

Share this project:

Updates

Submission history