DreamKnock turns your Apple Watch into a lucid dreaming trainer. While you sleep, it senses REM from heart-rate dynamics and wrist stillness, then plays a gentle haptic rhythm — quiet enough to slip into the dream instead of ending it. Train the rhythm by day with reality checks, so at night it means one thing: this is a dream. Wake to a map of your night: REM windows, cues, trends, dream journal. 100% on-device — no servers, no accounts. PH exclusive: 3 months free (see promo code) 🌙
No reviews yetBe the first to leave a review for DreamKnock
Maker
📌
Hey Product Hunt 👋
I'm the solo dev behind DreamKnock. Quick backstory: I've always been fascinated by the old sleep-lab experiments where researchers flashed lights at dreamers during REM and the light leaked into the dream — sometimes as a flicker, sometimes as something the dreamer recognized and used to realize they were dreaming. I tried an LED sleep mask once and lasted exactly two nights (I can't sleep with something on my face).
Then I realized the Apple Watch I already sleep with has everything needed to do the same trick without a mask: it can read heart rate, sense wrist stillness (your body goes nearly paralyzed in REM), and vibrate — silently. So DreamKnock infers likely REM from that signature and taps out a soft, distinctive rhythm into it. You train the same rhythm during the day with reality checks, so at night it means one thing to your dreaming brain: this is a dream.
Everything runs on-device — no servers, no accounts, nothing leaves the watch. 7-day free trial so you can actually test a full week of nights before paying anything.
Two questions I'd love this community's take on:
If you've tried any cue-based induction (masks, sound, apps) — what made you stick with it or abandon it?
Would you want the cue timing to adapt automatically (heavier late in the night when REM is longer), or would you rather control it manually?
Happy to answer anything about how the detection works, the App Review process (fighting Apple's toolchain was its own adventure), or the sleep science behind it. 🌙
Report
Apple Watch user here, this looks genuinely useful. One thing I'd love: let me pick the haptic pattern myself, maybe a custom tap sequence I design in the app. That way the cue feels more personal and less like a generic notification I might half-dismiss in my sleep.
Report
Maker
@derinalmargwoo Hey Derin, thanks for this — really good point.
Right now DreamKnock ships with three preset rhythms (Classic, Heartbeat, Waves) that you can pick between, but you're right that a fully custom sequence would make the cue feel like yours specifically, which matters a lot for something you need to recognize half-asleep. A pattern you designed yourself is inherently more distinct than any preset I could ship.
This is genuinely on my radar for a future update — a simple in-app tap editor (record your own rhythm, save it, use it as your nightly cue). No promises on timing yet since I'm one person building this, but I'm bookmarking your comment as the spec for it. If/when I ship it, I'll come back and tag you here.
In the meantime, if you try the preset rhythms and have a preference on tempo/spacing, that feedback would actually help shape what the default options sound like too. 🌙
Report
honestly the idea of a haptic cue that doesn't wake you is so cool. one thing though, would be great to export the dream journal as plain text or pdf. some of us like to move journals between apps or back them up outside the watch.
Report
the gentle haptics feel like a nudge rather than a jolt, and i actually stayed asleep through them. waking up to a clean map of my REM windows makes the whole thing feel worth it.
Report
honestly this looks really cool and the on-device approach is great. one thing i'd love to see is a sleep talk or whisper mic option for capturing mumbled dream snippets right after a REM window, since typing them out in the morning usually wipes the memory anyway.
Report
the haptic rhythm thing is genuinely clever, using one signal for two different contexts and letting your brain learn the association through the daytime training sessions. really thoughtful execution honestly
Apple Watch user here, this looks genuinely useful. One thing I'd love: let me pick the haptic pattern myself, maybe a custom tap sequence I design in the app. That way the cue feels more personal and less like a generic notification I might half-dismiss in my sleep.
@derinalmargwoo Hey Derin, thanks for this — really good point.
Right now DreamKnock ships with three preset rhythms (Classic, Heartbeat, Waves) that you can pick between, but you're right that a fully custom sequence would make the cue feel like yours specifically, which matters a lot for something you need to recognize half-asleep. A pattern you designed yourself is inherently more distinct than any preset I could ship.
This is genuinely on my radar for a future update — a simple in-app tap editor (record your own rhythm, save it, use it as your nightly cue). No promises on timing yet since I'm one person building this, but I'm bookmarking your comment as the spec for it. If/when I ship it, I'll come back and tag you here.
In the meantime, if you try the preset rhythms and have a preference on tempo/spacing, that feedback would actually help shape what the default options sound like too. 🌙
honestly the idea of a haptic cue that doesn't wake you is so cool. one thing though, would be great to export the dream journal as plain text or pdf. some of us like to move journals between apps or back them up outside the watch.
the gentle haptics feel like a nudge rather than a jolt, and i actually stayed asleep through them. waking up to a clean map of my REM windows makes the whole thing feel worth it.
honestly this looks really cool and the on-device approach is great. one thing i'd love to see is a sleep talk or whisper mic option for capturing mumbled dream snippets right after a REM window, since typing them out in the morning usually wipes the memory anyway.
the haptic rhythm thing is genuinely clever, using one signal for two different contexts and letting your brain learn the association through the daytime training sessions. really thoughtful execution honestly