❓ Problem

New parenthood is the most information-hungry period of an adult's life, and it happens at exactly the moment the brain stops working. Seven in ten parents lose around three hours of sleep a night in the first year, which adds up to 133 nights of lost sleep before the first birthday. Owlet's State of Parenting report found exhaustion is the defining emotion of year one at 61%, followed by overwhelmed at 48% and anxious at 32%, with 86% waking up to eight times a nightjust to check the baby is still breathing.

And they have already found the answer machine: a 2024 study found 52.7% of parents had used ChatGPT for parenting strategies. Sam Altman is one of them. The behaviour is not a hypothesis, it is already the default. The problem is the interface. Typing a paragraph into a chat box requires two hands, a clear head and a lit screen. At 3am a new parent has none of those, and a general-purpose model has no idea who they are, what they tried on Tuesday or how old their baby is in weeks.

Meanwhile the apps that do know the baby do not talk back in any useful way. Huckleberry has served more than 5 million families and raised $16m doing it, but it is fundamentally a logging tool with charts on top. It records that the baby woke at 2:40. It does not know your mother-in-law arrives Thursday, that you have already tried the swaddle, or that you are about to cry.

The parents who do get a second brain buy one. Overnight newborn care runs $300 to $600 a night in most US metros, with families spending $10,000 to $35,000 over an eight to sixteen week stretch. That is the true price of the job to be done, and roughly nobody can pay it.

✅ Solution

A voice-first AI companion built for the first 100 days of parenthood. Not a tracker with a chatbot bolted on.

• Voice as the primary input, not a feature. You talk to it while feeding, pacing the hallway, or driving the baby around the block for the fourth time. No typing, no screen, no waking anyone.
• It logs the boring stuff by listening. "She went down at ten past two and only took twenty minutes" becomes a feed and sleep entry without a single tap. The tracking data everyone else asks parents to enter manually becomes exhaust from a conversation they were having anyway.
• Memory is the moat. It holds the shape of your specific household: that naps after 3pm ruin bedtime, that you are combination feeding, that your partner does the 11pm shift, that you have already tried the white noise. Every answer starts from context instead of from scratch.
• It turns the chaos into a plan. Each morning it gives you one calm page: what happened overnight, what today probably looks like, one thing worth trying, and one thing to stop worrying about.
• Strictly non-medical, by design. It handles routine, logistics, reassurance and "is this normal" reframing, and it routes anything clinical to a human fast. The escalation path is the product, not the disclaimer.

Then it expands. First 100 days becomes first two years. Feeding and sleep becomes weaning, nursery transitions and toddler tantrums. The wedge is the hardest 14 weeks of a parent's life. The platform underneath is the family operating system.

📊 Key Numbers

Market size

• The global parenting app market was worth around $1.94bn in 2025 and is forecast to hit $6.02bn by 2035 at a 12% CAGR. That is the visible category, and it is understated, because it excludes what parents spend on human support.
• The US alone had 3,606,400 births in 2025. Every one of them is a fresh, high-intent, extremely motivated customer with a hard start date you can predict nine months in advance.
• If you charged $25/month across a nine-month subscription window, the US new-parent cohort alone is roughly an $800m annual pool. Add the UK, Canada, Australia and Ireland and you are comfortably past $1bn without leaving English.