Log food in ten seconds: search, quick picks, barcode

Food logging fails for one reason: it takes too long. If logging a roti takes a minute, you stop by Thursday. I spent more time on the Food tab than on any other screen because the goal was ten seconds for a normal meal. Here is how to get there.

The ten-second path

1. Open the Food tab.
2. Type the first few letters. "rot" shows rotli and roti, "dud" shows dudh and milk.
3. Tap the food. A portion stepper opens with live macros, so you see protein change as you go from 2 to 3 rotis.
4. Pick the meal segment (breakfast, lunch, snack, dinner) if the app guessed wrong from the time.
5. Tap Add.

That is it. Today's log below is grouped by meal, and each entry has edit and delete, so a wrong portion is a two-tap fix.

Search knows the names you actually use

The database has 235 built-in Indian foods, with Gujarati dishes included because that is where I am from and where the first users were. Thepla, khichdi, dhokla, undhiyu, bajra rotla, sev tameta: all there. Aliases cover the words people type in a hurry, so "rotli" finds the same item as "roti" and "dudh" finds milk.

Example search: type "dal". You get dal (tadka), dal fry, moong dal, chana dal, dal dhokli and sambar. Pick moong dal, set the portion to 1 katori, and the stepper shows roughly how much protein that bowl gives you before you add it. Most people are surprised by how little it is, which is exactly the point of seeing the number.

Category chips above the results narrow things down: grains, dals, vegetables, dairy, snacks, drinks and so on.

Quick picks and frequent foods

After a few days the Food tab starts showing your frequent foods at the top. If you eat the same breakfast five days a week, it becomes a one-tap add. Quick picks sit next to them with common items for the current meal time.

For a whole repeated meal, use Saved meals. Log your usual lunch once (say 3 rotli, 1 katori dal, 1 katori sabzi, 1 bowl curd), then save it as "office lunch". Next time it is one tap and the whole plate is logged.

Barcode scanning for packaged food

For packets, protein bars, milk cartons and the like, tap the barcode icon. The camera opens, you point it at the barcode, and the product is looked up on Open Food Facts, a public database. If it is found, the macros fill in and you just set the quantity.

Two honest limits. First, the live camera only works when the site is served over HTTPS or on localhost, because browsers block camera access otherwise. If the scan button does nothing, that is why. You can still type the barcode number by hand, or take a photo of the barcode and the app reads it from the image; both work anywhere.

Second, Open Food Facts is good for big brands and patchy for small Indian ones. When a product is not found, add it once as a custom food (see below) and it is yours forever.

AI photo estimate as a backup

Sometimes you are at a wedding and there is no way to log a thali item by item. Take a photo and the AI gives a rough estimate of what is on the plate and the macros. I want to be clear about what "rough" means: it cannot see oil, it guesses portion sizes, and it will sometimes call a dhokla a khaman. Use it when the alternative is logging nothing, then edit the entry if you know better.

Custom foods and the AI meal plan

Your mother's recipe is not in any database. Tap My foods, enter a name and the macros per serving (from the packet, or estimated once with the Coach tab), and it shows up in search like any built-in item.

If you do not know what to eat to hit your protein target, tap AI meal plan. It builds a day of meals around your targets and your diet setting, vegetarian included, and each meal has a one-tap add. Treat it as a starting draft, not a prescription.

How much to log

You do not have to log everything to benefit. My advice for the first month: log protein-heavy foods carefully and estimate the rest. The weekly score only checks whether you hit protein and whether you logged at all, so a quick honest log beats a perfect one you abandon.