GitFit.AI
Track any nutrient, habit, or activity daily with AI
126 followers
Track any nutrient, habit, or activity daily with AI
126 followers
Snap meal pics and describe, AI only counts nutrients, exercises, and activities you care about daily, with clean shareable contribution charts. Ask AI any question about your data, and design real-time, dynamic dashboard widgets to crunch numbers, displaying stats that matter to you, however you want.
GitFit.AI
@aaron_frost What do you think is your biggest competitive edge to NutriCam or or Kalo?
Wordwand
@aaron_frost "Git Hub commit charts for everything you care about" - that's the clearest pitch for a tracking app I've heard. Immediately understood the concept.
What stands out to me is the philosophy behind it. Most fitness apps assume they know better than you - here's your meal plan, here's your workout split, here's your macro target. But anyone who's serious about self-improvement already knows what they want to track. They just need a system that gets out of the way and lets them do it. The fact that you built this for yourself first and used it for 4 weeks before launching shows.
Genuine question on the grid system: how does it handle tracking things that are hard to quantify? Like sleep quality or energy levels throughout the day - is that where the mood/range type comes in? And can you set up grids that cross-reference each other, like "show me how my mood correlates with days I hit my protein goal"?
GitFit.AI
@diegodau
Hi Diego, I'm happy you see the value here!
To answer your question:
Energy levels throughout the day isn't supported, as each log would override the previous for that day / grid square.
Sleep quality would be something you can easily track.
The "Range" is actually just a coloring option for the grid. You can go green to red, or red to green, or just a solid gradient + optional turning red past a threshold (ex. too many calories), depending on your needs.
Yes indeed: there are two ways to do this.
"Compare graphs" button at the bottom of your dashboard shows you an overlayed version of any two grids of yours (or even others too if you want to compare)
By creating a custom dashboard widget with AI. Just one prompt, and you have one pinned to your dashboard.
Although I think what you would ideally want is a built in grid that does the correlation, so you can have it pinned to your dashboard.
Will also note that with gitfit.ai, you can ask any question about your data, and get an answer.
Thanks for the great questions!
Smart use of ai, guys)
I've a question.
How does it handle homemade meals vs restaurant food? And can it estimate portion sizes accurately from photos, or
do you need to input weights manually?
GitFit.AI
@denious
Hi Denis, thanks!
gitfit.ai handles photo logging the nutrients for both homemade and restaurant meals.
It can accurately estimate portion sizes from photos, with a <6kcal long term bias.
In any case, but especially for restaurant meals, gitfit can actually search the internet in real time to retrieve the restaurant nutrition facts, and apply those values, citing the sources for you to verify if you want.
Afterwards, you are given the option to fix the values for each nutrient you wanted to track manually or by asking AI to fix it however you need.
As someone who really cares about health and nutrition, I find this app super useful especially how it tracks calories, workouts, and more @aaron_frost I’m curious though: how does it estimate calories just from a photo, and how accurate is it? Also, are wearable integrations on the roadmap like auto-syncing activity data from Apple Watch, Garmin, or Fitbit so users don’t have to manually log every run?
GitFit.AI
@clairedo_04
While I still need to do more testing, on a sample of 100 randomly selected dishes, it scored <6 kcal bias, meaning that on average, long term, it is accurate enough for calorie tracking (assuming the dataset is somewhat reflective of your average diet and logging behavior). The analysis I did is at https://gitfit.ai/proof
And Apple Health is already integrated (steps, sleep), I have yet to add support for more variables but I could add in the rest in less than a day. I'd also love to add support for Garmin / Fitbit using their APIs. Wouldn't be difficult. Already support GitHub commit syncing too :)
Tracking macros is a pain. Having to enter my meals and snacks becomes a pain point and time suck. Eventually I just give up on tracking and just wing it. I'm interested to see how this addresses my pain point.
As someone actively tracking my own fitness journey, this resonates a lot! The idea of snapping meal pics and having AI count only the nutrients I care about is brilliant — most fitness apps overwhelm you with data you do not need. The shareable contribution charts are a nice touch for accountability too. Congrats on making the leap from Blizzard to build this!