I've been obsessed with a MIT paper about directing dreams. So I built an app around it.
A few months ago I fell down a rabbit hole of sleep science and honestly haven't climbed out.
It started with a 2020 paper from MIT's DREAM Lab showing that audio cues delivered during the exact moment you fall asleep measurably influence what you dream about. Not vaguely. Measurably. They called it Targeted Dream Incubation.
Then I found a Nature paper from 2004 showing people are 3x more likely to solve a problem after sleeping on it. Not because of rest. Because of how sleep actually restructures memory for insight while you're out.
Then a Harvard researcher in 2001 got subjects to focus on a specific problem before bed and 70% reported dreams that directly addressed it.
And I kept thinking: why is nobody building anything around this?
Sleep apps optimise for tracking. Meditation apps help you relax. Nobody is treating the 8 hours of unconscious processing you do every single night as something you could actually point in a direction.
So I started building Dream Director.
How it works
In the evening you pick an intention. A creative block, a hard decision, something emotional, athletic performance, whatever feels alive for you right now. You answer 3 questions. The app generates a personalised ritual using your own words and metaphors rather than a generic template. Binaural beats and ambient sound run all night.
In the morning a prompt fires before you touch your phone. 30 second voice log. Dream Director transcribes it and produces a Morning Insight Card with 3 observations tied to your intention.
Over time it builds a Dream Language Profile, basically a personal symbol dictionary built from your own sessions rather than some generic dream dictionary someone else wrote.
What building it has actually been like
The hardest part technically wasn't the generation. It was the structure. A ritual that works has four distinct phases with specific psychological functions and getting a language model to honour that structure while still feeling personal took a lot of iteration.
Morning insight is harder than the evening ritual. You're working with fragments. A colour, a feeling, a face. The failure mode is producing something that sounds meaningful but says nothing. I'm still working on this.
The Dream Language Profile is the piece I'm most excited about and least satisfied with. Tracking recurring symbols across weeks of logs and separating genuine patterns from noise is a hard problem and I haven't cracked it properly yet.
Where things stand
Landing page is live at dreamdirector.app. Waitlist is open. Founding members get lifetime 50% off whenever the app ships.
I'm a solo founder, pre-revenue, still building.
Not posting to pitch. Posting because I think deliberately using sleep as a creative and cognitive tool is a genuinely underexplored space and I want to find people who are already thinking about it.
If you've experimented with any of this, pre-sleep priming, dream journalling, anything adjacent, I'd really like to hear what you found.
This is fascinating, love how you’ve tied serious research into something actionable. The way you’ve structured the ritual and the Morning Insight Card sounds like it actually respects how the brain processes sleep rather than just being a fancy tracking app.
I’ve dabbled in pre-sleep priming and dream journaling a bit, and the recurring theme I noticed is that even subtle cues—questions, words, or sounds—can drastically change how you remember and interpret dreams. But the hardest part is always separating meaningful patterns from noise, which it sounds like your Dream Language Profile is trying to solve.
Curious: how do you handle false positives in the symbols—like when a recurring shape or feeling is just random vs. genuinely tied to someone’s intention?
Also, huge respect for doing this solo. Pre-revenue, research-heavy, and iterative—sounds like a lot of moving pieces, but super compelling space.
Hello, You enter before you sleep a couple of questions and the AI looks into it and ties it back to your dreams and builds a database. if recurring things are random it will most likely happen across multiple feelings. so if you said you were sad or happy at the beginning of the ritual and that recurring thing happends but it also happends when you are in distress or depressed its most likely random. it will look into all of that. and if people find that overtime the app is wrong i can always look into the data. but i dont have users yet so i dont know for sure if that is going to work.
Thanks for the question!
This all sounds very interesting, the only question I have is why. Why to "direct" dreams? Another problem is do you know that we don't remember most of dreams?
Directing your dreams means to prime your dreams for bigger things than sleeping. you direct your dreams and subconcious thoughts to make you feel better/less sad. To overcome a creative block or something else. (all backed by science) And the dont remember the dream part. Most people know their dream right when they wake up. The app asks immidiate when you wake up to talk into the app what youve dreamed. (or type it in, whichever you want) And if you still cant remember the dream you can skip the dream part and only use the pre sleep part for that day.
Thanks for the questions!
Well, yes, people can remember their dreams after wakeup. But it's just a small part of the sleep. The whole sleep circle has several phases when people see dreams and they remember only the last one. About subconscious - it looks like a "grey" area, I mean you never know what exactly makes you sad or cheerful, so without this knowledge, the direction could cause the opposite result.
For example, I remember it happens to me several times in the past - when I was in a specific place (it was just a regular supermarket) I felt very strong depressing feelings and I still don't understand why. Does it mean that I can "direct" to see less supermarkets in my dreams? Or more? And how can it affect my day time mood?