A personal starting point
Crmbls AI started from a very personal place.
I live with several chronic health conditions, including FSHD, and I also have fructose intolerance. Because of that, food has always been something I had to think about more consciously than many other people.
Over time, I learned that simply "eating healthy" is often too vague. I wanted to understand what I actually consume during the day. Not only calories, but also nutrients that may matter for my energy, muscle health, hydration and overall wellbeing.
But tracking all of that manually was difficult. The information was scattered, the process was time-consuming, and the data was not easy to connect in a useful way.
The spreadsheet phase
At the beginning of the year, I tried to solve this with a spreadsheet.
I started tracking values that felt important to me, especially around muscle health. I looked at how much protein I was getting from normal food, how much I might need to add, and how my daily intake looked when I included essential amino acids.
For a while, it helped.
But it also showed me the problem very clearly: doing this every day takes a lot of time. After a few weeks, I became inconsistent and eventually stopped tracking regularly. Not because it was useless, but because it was too much work to keep up with.
AI, a turning point?
Around that time, I had also started experimenting with coding AI.
A little later, after I had mostly stopped using my spreadsheet, I saw a video about an app that used AI to track food from photos. I tried it immediately and I was impressed.
The idea felt simple and powerful: take a photo, let AI help, and avoid the daily work of calculating everything yourself.
But it also raised one important question for me:
How reliable is this?
I realized that many nutrition values were estimated. Sometimes that can work well, but sometimes it can also be far off. For what I personally wanted, that was not enough. I needed more transparency and more detail.
Yes, AI was my turning point!
I did not continue using that app, but I kept the "WOW" moment.
For almost four months now, I have been building my own AI-assisted tracking app. But I did not want it to feel like pure magic, where AI looks at a photo and simply gives a result.
I wanted to know where the nutrition data comes from.
That became one of the core ideas behind Crmbls AI: use AI as a helper for food recognition, but combine it with structured food data, reviewable results and clear reliability signals.
This system changed several times while I was building it. I rebuilt parts of it whenever it did not feel precise or transparent enough. Now, for the first time, it feels like something I am genuinely happy to use myself.
Crmbls AI, my daily game-changer
Crmbls AI has become more than a food scanner.
It helps me track meals, hydration, supplements, goals, wellbeing and daily logs in one place. It also helps me look beyond basic calories and macros.
Because the app can track deeper nutrient values, I started thinking about how it could become even more useful in daily life. Everyone is different, and everyone has different priorities. For me, nutrients such as calcium, magnesium, omega-3 and protein-related values can be especially interesting.
That led to the idea of a Lifestyle Score.
The goal is not to diagnose anything or tell people what they must do. It is a non-medical orientation based on the values someone chooses to focus on. If I set goals for certain nutrients or habits, the app can help me see how consistently I reach them.
For me personally, this has become a real game-changer. It motivates me to stay more consistent and gives me a clearer picture of my day.
It will not cure my conditions, and it is not a medical tool. But when there are limited treatment options, I want to use every daily lever I can influence and nutrition may be one of the most important ones.
Beta phase
Because Crmbls AI has already become useful for me, I would now like to share it with others.
The app is currently in beta, and I am looking for Android users who want to test it in real life. Positive feedback is welcome, but honest and critical feedback is even more important.
I have also created a Reddit community where beta testers can share thoughts openly, report issues, suggest improvements and help shape the app.
The beta phase is expected to run for around 12 to 20 days. During this time, the app needs real testing before it can move toward a public Google Play release.
My hope is to keep Crmbls AI as accessible as possible. I would love to offer a long-term free version, especially for trying the app or using simple daily tracking features like water logging. But because AI scans and backend processing create real costs, I cannot promise the final model yet.
For now, the goal is simple:
Share the app, learn from real users, improve it step by step and build the tool I originally wished I had for myself.
Future
Crmbls AI is still evolving.
My goal is not to make the app feel as fast or as magical as possible. My goal is to make it useful, understandable and trustworthy. If I track something, I want to have a clear sense of where the data comes from, how reliable it may be, and whether it actually helps me make better daily decisions.
That is why honest feedback matters so much during beta. I want to hear what feels helpful, what feels confusing, what takes too much effort, and where the app should become simpler. Positive feedback is great, but critical feedback is even more valuable.
One area I especially want to improve is the Lifestyle Score. It is not a medical score and it is not meant to judge anyone's health. It is a personal orientation based on the goals and focus areas someone chooses.
At the beginning, the score can be low or incomplete because the app does not have enough data yet. As more days are logged, the picture becomes clearer. Over time, the score should help show whether someone is becoming more consistent with their own selected goals. For example compared with an earlier period.
That is the direction I want to keep building toward: a calm, personal system that helps people understand their own routines better, track what matters to them, and improve step by step.
Crmbls AI will not be finished with the first beta. The beta is the beginning of making it better.
