Anfold lets you log food and training in plain words and turns them into entries with ingredients, macros, micronutrients, and per‑muscle training load. It matches meals to official nutrition databases, labels estimates, uses scientifically grounded methods, and shows the assumptions behind the numbers. Use day and week views to see what changed, spot gaps, and plan next steps. Ask follow-up questions with full context, adjust drafts, and build meal plans around the day you already had.

Hey PH 👋

I'm Vlad, developer of Anfold.

What kept bothering me, aside from tracking taking more work than it should, was that even after logging meals, workouts, notes across separate apps, I often still had to piece things together, figure out what mattered and what to do next on my own.

Anfold is my attempt at making that loop easier and more complete: describe what happened in one place, get entries with the numbers worked out, use the data to understand what mattered and then plan next steps.

What it does:

Log meals and workouts in plain language. You describe what you ate or how you trained. Anfold turns it into reviewable entries with the numbers worked out.
Understand your day and week. Built-in daily and weekly views, reports, and assistant questions show what changed, what mattered, and where the gaps are.
Plan what comes next. Get meal suggestions and nutrition plans for the days ahead.

A few things that might catch your eye:

The AI is not asked to do arithmetic. The numbers come from deterministic math, not LLM guesses: USDA data plus cooking-state transforms for food, established training-load formulas for exercise. The full method is at anfold.app/approach.
The assistant answers from your real logged values. "What's my top sodium source this week?" returns an actual number pulled from your data.
Meal plans are built for you specifically. They look at the days you've already had, anything you've planned ahead (pizza with friends tomorrow?), and they're rooted in your profile: preferences, goals, allergies.
Anfold uses verified USDA data where it can, deterministic transformations where needed, and labeled estimates where the data gets fuzzy. The point isn't fake precision, but useful estimates with the seams visible.

What's still early:

A principle behind Anfold is that nutrition, exercise, and sleep are connected and affect each other meaningfully, handling only one leaves you partly blind. Today nutrition is fully there, exercise works but has room to get better, and sleep isn't in yet, it's on the way (you can still track through notes).
No wearable sync yet.
Meal plans are currently capped at one day. Exercise plans aren't in yet, but they're on the way.

If you've tracked food, training, or sleep before: where did the process break down for you?

Live on iOS and Android, bilingual EN/DE. Free tier covers form-based tracking. The AI parts run on a credit system starting around 5€/month; about 10€/month should cover an average month of logging and chats for both food and exercise.

Thanks for taking a look, appreciate any feedback.

— Vlad

Berlin, Germany