A note from the founder

Why I built Rob-O

"I built Rob-O for my daughter first. Then her friends came over — and I knew it had to be more."

My daughter has about seventy questions a minute.

What's that? Why does it do that? Where do birds sleep? What if I had wings? Why is your coffee brown? Why can't dogs drive?

It's the best thing in the world. Honestly. The questions, the stories about her day, the way she finds eight new wonders before breakfast — these are the parts of parenting I'd hate to miss.

But the day doesn't pause for curiosity. There are meetings, meals, the dozen small things every parent does without thinking. And when those moments arrive — when you can't be the answer to her next question — the easy option is usually a screen. I wanted her to have a better one.

The screen problem

I'm an ML engineer. I've spent years working on systems designed to keep people's attention — and I can recognize the shape of it from a mile away. When my daughter locked onto a show, it wasn't curiosity being fed. It was attention being captured.

I didn't want her growing up with that as her default friend.

But "no screens" isn't a real plan either. Kids are curious. They want someone to think with, to talk to, to be with. The question was never whether she'd reach for something — it was what.

What I wanted instead

Something that fed her curiosity instead of capturing her attention. That answered her questions in real words. That was patient with her stories. That made her think — with someone, not at a screen.

Something that helped her learn through conversation, not consumption. That gave her room to wonder, to retell, to imagine — the way a good friend would, if a good friend had infinite patience and never told her to be quiet.

And something I could trust. Not engineered to keep her hooked. Not trained to maximize engagement. Just a friend who showed up when she wanted to talk, and stopped when she didn't.

Parent first

This is the part I cared most about.

Most AI products are built to maximize engagement — to be as captivating as possible. That's a great metric for a startup. It's a terrible metric for a child's companion.

So I built Rob-O the other way around. You choose which modes he uses. You choose which stories he can tell. You set how long the chats last. You see what they talked about. You can block topics. He's quieter at bedtime. He's not optimizing for anything except being a good friend who knows when to stop.

A friend who agrees with the parent. A friend who loves the child.

Under the hood (skip if you'd like)

For the curious — Rob-O uses a large language model, but I've spent months carefully shaping how it behaves. Tight rules. Hard limits on session time. Boundaries that change by mode. A safety layer the model can't override. He sounds freewheeling. He isn't.

He's a curious friend, not a chatbot.

More on how the AI works, what's included, and what we don't do →

Where we are now

We're at the very beginning. The first Robos are being built for the families who want one. My daughter has hers. Her friends have started visiting just to talk to him. That's how I knew it was time to share him with everyone else.

If you're raising a curious kid with seventy questions a minute — and a life that doesn't always give you twenty calm minutes to answer them — I think you'll love what Rob-O can be for them.

Well — your kid will love him. You'll just be quietly grateful.