One AI agent for your whole family — one that helps with homework, manages the family calendar, and connects to the tools you already use. With per-child profiles, content guardrails, no training on your data, and full transparency about what the AI sees and does.
The big chatbots are powerful — and built for generic adults. Letting a kid use them often means letting a model see things you'd rather keep private, answer questions you'd rather it didn't, or train on your family's data.
Your conversations don't train our models. PII gets redacted before any external service ever sees it.
Per-child age settings adjust tone, vocabulary, and content categories. Content filters that actually understand context.
Safety events surface to the parent dashboard. No surprise overreach, no hidden conversations.
We didn't water it down for families. The guardrails, redaction, and capability gates security teams use at work? Your family gets the same stack — on every plan.
Every prompt your child sends gets checked for unsafe content, prompt injection, and policy violations before the AI ever processes it.
Every link the AI considers visiting is scanned. Phishing, malware, and unsafe content blocked before reaching your kid.
Names, addresses, phone numbers, emails — redacted with placeholders before any external tool sees them. Always. Even from the AI itself.
Decide which tools each child can use. Email send disabled for the 8-year-old? Done. Calendar read-only? Done.
Set up a profile for every member of your household. Each one gets the right tone, the right tools, and the right boundaries — automatically.
We don't train on your conversations. Not your family's, not anyone's. Your data is yours.
PII is redacted before tools see it. When the agent uses an outside service, names and identifiers are replaced with placeholders. We log what was redacted — never the original.
Every action is logged for you. Parents can review what the AI did, what it considered, and what the safety stack caught. No black boxes.
Built by veterans. The same operational discipline we'd bring to a forward operating base, applied to the AI in your kitchen.
Per-child profiles support age bands (5–8, 9–12, 13–17, and adult). Each band adjusts content categories, tone, and vocabulary automatically. You can layer additional restrictions on top — tool capabilities, time-of-day rules, and content category overrides — per child.
Yes. The parent dashboard shows every conversation, every tool the AI considered using, every guardrail event, and every PII redaction — per child, with timestamps. No hidden activity, no surprise behavior.
Not unless you explicitly enable a write capability for that child. The default is read-only across the board. If the AI thinks a write action would help, it drafts the message and surfaces it for parent review instead of acting on its own.
The input guardrails catch it before the AI processes the request — a three-stage pipeline: a fast heuristic filter, then a parallel LLM guardrail + URL safety check, then a block or rewrite. The child sees a kid-appropriate redirect; the safety event lands in the parent dashboard.
Yes. The same input guardrails, PII redaction, capability gating, and audit logs run on every plan. Safety isn't a feature we gate behind upgrades.
Yes. Per-child profiles are independent — different age bands, different tool capabilities, different content restrictions, different time-of-day rules. A 16-year-old can have a different ruleset than a 9-year-old, and both differ from a parent's profile.
Free to start. $5 in credits on us. No credit card required.