metrIQ started as a spreadsheet for my own training. The numbers were fine, but a spreadsheet never tells you anything — it just holds what you typed.
So I switched to asking ChatGPT. It could explain things, but it forgot my history, answered the same question differently on different days, couldn't give me the stats I wanted, and buried it all in generic advice. I wanted a coach who remembered me.
So I built my own. The first version was a rough thing in Lovable, mostly because I wanted an excuse to try it. Then friends started using it, it actually helped them, and I rebuilt the whole app properly for launch.
That's metrIQ. It's Sophia, your coach. A sentence is enough to log something, and unlike a chatbot she remembers your goals, your weeks and what went sideways last month. Her targets and nudges come out of that. She doesn't only track your fitness journey, but offers insights, advice and support.
I am eager for people to try it and I appreciate all feedback, warm and cold, especially from people who are training or people who are in a weight loss journey. If you tried it and didn't like it, drop me a line about why, maybe i can fix that
Nikos
I'll just never feel the same about "insights, advice and support" coming from a LLM vs. a real person. I know that perhaps in a blind A/B testing via text channels I probably won't be able to tell the difference, at least not right off the bat, but there's just something profoundly wrong with humanizing this. Personally I'd very much like an "AI training framework" (tried many but no fit yet), but I don't need a digital cheerleader.
RE: calorie counting, I used to work on a calorie counting app that worked on delivery meals (back in China - nobody cooks there). It was way back then, so we talked to the chefs and had recipes w/USDA-like DBs to estimate calories. The assumption was that you measure your intake from the app and your weight, if it doesn't go up/down as expected, you just adjust a coefficient every N period, i.e. not losing weight? ok let's set all calorie estimates at 1.1x for you. Rinse and repeat until you're losing weight.
Thank you for the reply, I appreciate the feedback
I understand you oppose the very concept of the app, but I would still like to ask: is there something that would actually change your mind? A change in position, a different way of using the LLM, something else?
I am sure there will be lots of people feeling the same, so I am wondering if there is any way to mitigate the problem or I should accept that those users will never use the app
(thank you about the calorie counting part as well, that is truly useful for me)
I used to body build and I just don't understand how these AI calorie tracking tools are actually useful.
For example, 1 tablespoon of oil is 120 calories and 14g of fat. Miscalculating 2 tablespoons per day for a month is 3600 calories (or an 1.5 day's worth of food untracked). If you are trying to track what you eat, you end up cooking yourself and eating the same meals everyday so you aren't constantly re-measuring everything.
If your goal is to evaluate food for allergies, then tracking what you eat makes sense. But when it comes to macro tracking, how much is just as important as what.
I can see 2 ways where metrIQ can help
1) First of all, you sound like a knowledgable person. In your case, you already know what to eat. So for the tracking part, Sophia will offer you convenience mostly. But she can also detect patterns and provide insights. Things like "I notice that during the evenings you tend to overeat. How about you try XYZ to improve that". Or you could ask directly questions about your journey or even projections ("When do you think I will have reached 67kg based on my weight trend").
In general though, you are right: for someone knowledgable and disciplined enough to track meticulously and eat standard meals, a calorie tracker is less useful. The discussion and statistics part remains relevant though
2) For people we are not so advanced though, it helps both by helping them understand things better, but also flagging things: lets assume an inexperienced user underestimates what they are logging. After some time, Sophia will pick up that the trend is not moving towards the right direction and will make suggestions based on what has been logged: is the user not logging everything? is the user hitting their targets in the app and still gaining weight? are his logs wrong, is his estimated activity levels wrong etc
>For example, 1 tablespoon of oil is 120 calories and 14g of fat.
You should not use them, but for some foods they may work. So if you have a person who isn't going to track the traditional way, then tracking something is better than tracking nothing.
Let's say you are consistently undercounting calories. In that case, free app like MacroCodex (which uses adaptive TDEE tracking) can offset that error and estimate your maintenance calories helping you make better adjustments going forward.
Deficit = TDEE - Intake
macrocodex will simply lower your tdee estimate if you are undercounting, then you'll cut calories further and you'll end up in deficit and lose weight.
Compensation works only when the error is consistent.
It does not solve macro tracking. This doesn't know if the overage is from fat, carbs or protein.
Adjustments are very delayed and noisy. Your body weight fluctuates with water retention, based on plethora of variables like stress, sleep quality, hormone cycles, etc. The feedback loop tracking weightloss with inaccurate calorie and macro tracking delay or fully impede goals.
>Compensation works only when the error is consistent.
many people have predictable eating pattern.
>It does not solve macro tracking. This doesn't know if the overage is from fat, carbs or protein.
For purpose of weight gain or loss it doesn't matter.
Deficit = TDEE - Intake
>Adjustments are very delayed and noisy. Your body weight fluctuates with water retention, based on plethora of variables like stress, sleep quality, hormone cycles, etc. The feedback loop tracking weightloss with inaccurate calorie and macro tracking delay or fully impede goals.
well macrocodex uses an algorithm which is designed to deal with all these issues and produce accurate maintenance calorie estimate from calorie intake and weight log data, you can find more details here: https://macrocodex.app/knowledge/rethink/adaptive-tdee/
Later i plan to add more details about the algorithm in a dedicated page.