Tuck

Tuck · About

About Tuck

TL;DR. Tuck is built by Veronata, Inc. — a small US team making a baby monitor that uses two iPhones, works without Wi-Fi, and adds AI lullabies in a family member's voice. We're not building medical hardware; we're building thoughtful software that treats parents and babies with respect.

Why we built Tuck

Most baby monitors ask you to buy hardware that's worse than the iPhone you already own. The cameras are blurrier; the apps are slower; the parent units lose Wi-Fi the moment you walk into the basement. We had a drawer full of retired iPhones and an unanswered question: why isn't there a baby monitor that just uses these?

The answer turned out to be: building one well is hard. Real-time WebRTC across two devices is finicky. Cry detection that doesn't fire on every neighborhood dog needs careful on-device ML. And the single most important thing — that the monitor keeps working when the network doesn't — required us to build a custom Bluetooth Coded PHY transport, which is the kind of work that doesn't pay off on a sprint board but does pay off the night you're in a hotel room and the Wi-Fi is broken.

Tuck is what you get when a small team takes those problems seriously, ships the boring foundational stuff first, and only adds AI features after the core monitor works.

What we believe

The tech, briefly

Tuck is native iOS — Swift + SwiftUI, AVFoundation for media, Core Bluetooth for the offline transport. Real-time media uses LiveKit's WebRTC stack on the active path; cry detection runs on-device using a small Core ML model. AI features call out to Gemini 2.5 Flash for scene understanding and Mureka for lullaby generation, both with explicit consent. Convex acts as the state-broker for pairing and subscriber metadata — no audio or video passes through it.

We picked this stack because each piece has a clear failure mode we can communicate to a parent. "Wi-Fi dropped, falling back to Bluetooth." "Lullaby generation needs internet." "Cry detection runs on your phone, no internet needed." Tech that's hard to explain to a sleep-deprived parent is tech we don't ship.

Frequently asked questions

Who makes Tuck?

Tuck is built by Veronata, Inc., a US-based company. The team has prior experience shipping real-time WebRTC products, AI-powered consumer apps, and privacy-respecting iOS software. Tuck is our first dedicated baby-care product.

Why build a baby monitor as software, not hardware?

Most parents already own one or more iPhones. A modern iPhone has a better camera, better microphone, better processor, and better battery than 90% of the dedicated baby-monitor hardware on the market. Asking parents to buy another piece of hardware to do something their existing devices already do well felt wrong. The interesting problem is the software — making two iPhones into a real, reliable monitor — not the hardware.

What does "not a medical device" really mean?

Tuck does not diagnose, monitor, or treat any medical condition. We do not detect SIDS, sleep apnea, breathing irregularities, oxygen saturation, or heart rate. If you want medical-grade monitoring, look at FDA-cleared devices like Owlet Dream Sock — they're a different category of product with different regulatory clearance. Tuck is a connected baby monitor in the same category as Nanit or Cubo Ai: a great pair of eyes and ears, plus AI conveniences.

Why does Tuck use AI? Isn't that dystopian?

AI in Tuck does two things: it describes what the camera sees in plain English (so you can glance at your phone instead of staring at a video feed), and it composes lullabies. We're not running facial recognition, behavioral profiling, or sleep-pattern data sales. The AI is opt-in, requires explicit consent, and never leaves your phone in plaintext. Apple requires this consent flow for any third-party AI in baby-care apps; we agree with Apple.

What will Tuck never do?

We will never (1) make medical claims about SIDS or breathing detection, (2) sell or share your audio/video data with advertisers, (3) clone the voice of someone who isn't present and consenting, (4) hold your monitor footage hostage behind a subscription paywall — the Free tier is a real, complete monitor, forever, (5) ship a feature that requires the device to phone home to keep working.

How do you make money on the Free tier?

We don't, directly — Free is a loss leader. We make money from Pro and Pro+ subscriptions for the AI features (lullabies, scene understanding, voice cloning). The economics work because the marginal cost of a free Tuck user is tiny: no proprietary hardware, no cloud storage of nursery footage (we don't store it), and the LiveKit WebRTC is only spun up while monitoring is active. About 5% of usage is the active stream cost — the other 95% is essentially free for us, so we can afford to give it away.

Where is Tuck based?

Veronata, Inc. is a US company. Customer data is hosted in US data centers. We follow US privacy law (CCPA, COPPA where applicable for content involving minors), and we publish a plain-English privacy policy at tuck.baby/privacy.

When does Tuck launch?

We're targeting an iOS launch in 2026. Beta testers get free Pro for the first year. Sign up on the home page for early access — we'll email you when Tuck is ready.

Get in touch

Press: hello@tuck.baby
Support: tuck.baby/support
Privacy: tuck.baby/privacy

Try Tuck early →