Find the ingredient, not just the meal.
Snap meals, say how you feel; Trigger AI weighs each ingredient separately.
Free to log. Pro shows what it finds. No ads. Coming to iPhone.
How you felt after 13 meals
A dot: the ingredient was in that meal.
- Garlic Strong: in 8 of 9 meals you felt worse after, and 0 of 4 you felt fine after.
- Onion Can’t separate yet: in 5 of 9 meals you felt worse after, each time with garlic, and 0 of 4 you felt fine after.
- Basil Can’t separate yet: in 3 of 9 meals you felt worse after, each time with garlic, and 0 of 4 you felt fine after.
- Cheese Weak: in 2 of 9 meals you felt worse after, and 4 of 4 you felt fine after.
- Tomato Cleared: in 4 of 9 meals you felt worse after, each time with garlic, and 3 of 4 you felt fine after.
- Wheat Cleared: in 1 of 9 meals you felt worse after, each time with garlic, and 3 of 4 you felt fine after.
Illustrative — an example pattern, not a real result.
Garlic was in 8 of the 9 meals you felt worse after and in none of the 4 you felt fine after: a strong pattern. Onion and basil only ever came with garlic, while garlic also turned up in 2 bad meals without either, so neither can be separated from it yet. Tomato and wheat were eaten without garlic and you felt fine every time, so both are cleared. Cheese is mixed so far: in one bad meal with no garlic, but in all 4 fine ones.
- Scan: the app names the meal, tomato and basil pasta, and lists its six ingredients — tomato, onion, garlic, basil, wheat from the pasta, and olive oil. Garlic is already on your watchlist.
- Check in: after the pasta, the app asks how you feel. This time it is mild bloating, so the meal joins the ones you felt worse after. “Feeling fine” counts too: yesterday’s fine lunch moved every ingredient in it down the list of suspects.
- Pattern, an example of what Pro shows: garlic was in 8 of the 9 meals you felt worse after and in none of the 4 you felt fine after, a strong pattern. Onion and basil can’t be separated from garlic yet, and cheese is weak. Tomato and wheat are cleared.
- Protect: while you avoid garlic, the app flags a Caesar dressing that contains it, before the first bite. A grilled chicken salad checks out with no garlic found, and a barcode scan of oat crackers shows none listed.
Illustrative — an example pattern, not a real result.
Why the usual approach stalls
An elimination diet cuts several foods at once. If you feel better, you have learned that something in that group mattered — not which one. Adding them back one at a time puts you weeks in and still guessing.
A food diary records meals. But a reaction isn’t to a meal. It’s to something inside it: an ingredient, an additive, something that came along for the ride.
And the ingredients you eat most often travel together. Cheese arrives with bread. Neither one ever looks guilty, because neither one is ever alone.
How it works
Scan your meal
Take a photo, scan a barcode, or type it in. Trigger AI breaks the meal down to ingredients rather than meal names, because “pizza” is not a suspect — something inside it might be.
Say how you feel
About thirty seconds. “Feeling fine” counts just as much as a bad day, because that is how a suspect gets cleared.
Get the pattern
Trigger AI surfaces an ingredient when there is enough evidence to be honest about it, and not before. Until then it keeps watching.
Protect
If a pattern emerges, scan a meal before you eat. Trigger AI can flag a dish that contains that food.
We would rather be slow than wrong
~2weeks
The engine needs roughly two weeks of logging before it has enough data to look for patterns at all.
We could hand you a guess off a handful of meals. We won’t. A wrong trigger means avoiding a food for years for no reason — so when the evidence isn’t there yet, we’ll tell you it isn’t there yet.
Plan on about two weeks of logging before there is enough data to look for a pattern — sometimes longer.
Free to log. Pro shows what it finds.
- Meals
- Check-ins
- Post-meal prompts
Logging and check-ins are always free. When a pattern emerges, Pro unlocks it, along with more photo scans. If nothing stands out, you’ll see that for free.
Questions
Is this a medical diagnosis?
No. Trigger AI is not a medical device and it does not tell you what is wrong with you. It looks for statistical patterns between what you log and how you say you feel, then shows you what it found and how strong the evidence is. What you do with a pattern is a conversation for you and your doctor.
How much do I have to log before it finds anything?
Roughly two weeks of logging, before it has enough data to look for patterns at all. That is a floor, not a countdown: past it, an ingredient surfaces when the evidence separates it from everything it travels with, not on a schedule. Some never separate — they show up too rarely, or too constantly, to be told apart from everything around them.
What if I react to more than one thing?
That is part of why cutting foods out by hand is so hard. Trigger AI scores ingredients independently, so more than one can surface. It also means each one has to earn its own evidence, which takes more logging than a single clear match would.
Do I have to log every single meal?
Gaps weaken the analysis rather than ruin it. What matters more is that the meals you do log are complete — including the ones where nothing happened. A quiet meal you skip is exactly the meal that would have cleared an ingredient.
What happens to my data?
Your logs are yours. We do not sell them and we do not run advertising against them. The Privacy Policy sets out what is collected, who processes it, and how to have it deleted.
Is it free?
Free to log. Pro shows what it finds. Logging — meals, check-ins, and the post-meal prompts — is always free, because that is what the analysis runs on. When a pattern emerges, Pro unlocks it, along with more photo scans. If nothing stands out, you’ll see that for free. Pro’s price is not settled yet, and we would rather say nothing than name a number we then change.
What if the answer is wrong?
It can be. A pattern is evidence, not proof, so Trigger AI shows you what the match is based on instead of only the conclusion. It keeps updating as you log: when later meals contradict a pattern, the pattern weakens.
Does it work for digestive symptoms?
Trigger AI does not work on conditions. It works on what you record — you describe how you feel in your own words and on a simple scale, and it looks for ingredients that line up with those entries more often than chance would explain. It is not a substitute for medical care and it does not tell you what you have. It finds patterns between what you eat and how you feel, and what you do with one is between you and your doctor.
Get it the day it launches
One welcome email now, one when it launches. Nothing else.