We're parents who got tired of waiting.
Gamekeeper started in a kitchen in 2024, after Priya's 9-year-old got asked for her Snapchat handle by a 30-year-old stranger in a Roblox DM. Two years later, we protect 50,000 families. Here's who we are, and what we believe.
The kitchen-table version.
In March 2024, Priya — at the time leading T&S product at Discord — found a stranger in her daughter's Roblox inbox. The account was three days old. The conversation had been going on for a week. Roblox's own moderation hadn't caught it.
She called Marcus, who'd been at Meta building content moderation models. The same day, they realized: every parent they knew had a story like this, and the platforms — for all their announcements — weren't solving it at the rate kids were getting hurt.
Six months later, Gamekeeper shipped. Today we work with 50,000 families across 38 countries. We've flagged over 2.4 million risky interactions. Our research team contributes to ICAC task force data. And we still get a hand-written thank-you note from a parent at least once a week.
We're building this because we're terrified of the world our kids are growing up in — and because we're stubborn enough to think we can change a piece of it.
Four principles that shape every product decision.
We don't surveil. We protect.
Most monitoring tools read everything your kid says. We read what strangers say to your kid. The difference matters. Kid privacy is part of kid safety.
Alert fatigue is a safety failure.
If you stop reading our pings because we cried wolf, we failed. We tune aggressively. We'd rather miss a low-severity issue than train you to ignore us.
Kids should know we exist.
Hidden surveillance breaks trust the day it's discovered — and it's always discovered. We tell kids what we do, and we explain it in their language.
We work with researchers, not against platforms.
We share aggregated threat patterns with Roblox, Microsoft, and Epic's safety teams. We want them to win. We're a backstop, not a replacement.
The people you're trusting with your family.
Former product lead at Discord trust & safety. Mother of two. Quit her job after her 9-year-old got asked for a Snapchat handle in a Roblox DM.
Ex-Meta ML researcher who built content moderation models at scale. Father of three. Believes most kid-safety software treats kids like criminals.
Pediatric psychologist (PhD, Stanford). Co-authored the 2024 ICAC study on grooming patterns. Doesn't let her own kids on Roblox without Gamekeeper — that's not a slogan, it's a fact.
Built threat-detection models at Cloudflare. Trains our grooming-pattern classifier. Spent 8 months reading actual grooming transcripts to label our training data — so your kid doesn't have to live through one.
Start your 14-day free trial
No credit card. Cancel anytime. We only ping you about things that actually matter.
Start free trial