Your urge to smoke,
understood in real time.
Brume learns your risk windows — the coffee, the break, the evening — and steps in before the craving peaks, not after.
Analyzing
Stress: Moderate↘
Craving control
Improving
Preventive nudge
Breathe 3 min
Backed by clinical trials published in
- JAMA Network Open
- JMIR
- npj Digital Medicine
- Nicotine & Tobacco Research
- PLOS ONE
We know your moments.
The coffee, the break, the evening out. They aren't weaknesses — they're windows. Brume learns them and gets you ready before they arrive.
The first coffee
Coffee anchors the habit from the moment you wake. Brume preps you before the first sip.
The 3 p.m. slump
Energy dips, stress rises: your most predictable window.
Drinks with friends
Alcohol and company push the risk up. Brume anticipates it with you.
How it works.
We don't wait for the craving. We see it coming.
01 / 03Measure
A micro check-in measures the 6 validated relapse-risk variables. ²
02 / 03Predict
The engine crosses your answers with your clock — the strongest predictor. ³
03 / 03Intervene
A brief intervention before the peak: breathing, urge surfing. ²
A hybrid
architecture.
Clinical rules first, machine learning on top — to step in at the right moment, not at random.
EMA (Micro-measures)
Measure the real state, at the real moment.
The temporal prior
Time of day predicts better than GPS. Across 1,784 episodes, temporal features dominate.
What the science doesn't (yet) say: Fine-grained prediction in advance stays modest — so Brume keeps an SOS button always within reach. An app isn't a treatment: for replacement therapy, talk to a professional or a quitline.
Your risk climbs toward 6 pm. A 3-minute breathing break at 6:05 should be enough.
The science behind Brume.
Every design choice is backed by a clinical trial or a meta-analysis. Here's the evidence — including the parts that keep us humble.
Verified abstinence at 26 weeks: the adaptive app beats the standard one. ²
Staying active at 26 weeks multiplies 12-month quit odds by 5. ¹
Temporal features dominate prediction; location is marginal. ³
Geofencing: a positive but underpowered signal, hence our caution. ⁴
An avatar boosts content use — an engagement effect, not abstinence. ⁵
Forecasting "one hour ahead" with ML alone stays moderate. Brume owns it. ⁷
Rigor, not a promise.
Every part of Brume is backed by published data. Here's the order of magnitude.
They see the craving coming.
A small community is already testing Brume before launch. The theme that comes up: anticipate, without the guilt.
I thought I knew my triggers. Finding out my real peak is the 3 p.m. coffee changed how I get ahead of it.
Brume doesn't judge me. It warns me beforehand, calmly — exactly what I was missing.
A method grounded in science rather than willpower alone. That builds trust.
Imagine everything you get back.
Quitting isn't giving something up. It's the opposite.
Your breath
Stairs without a second thought. A deep breath, just because it feels good.
Taste
The first coffee that finally tastes like coffee. Meals become pleasures again.
~€4,500 / yr
A pack-a-day budget, back in your pocket. Month after month.
Questions? We've got answers.
Ready to breathe free?
Join the first people rewiring their habits with Brume. Join the program and start quitting at the right moment — not after the peak.
Free for the first cohort. iPhone first, Android next.