About Ania

The story

Ania started because of my grandfather. He's sharp, independent, and the exact person scammers hunt: he gets the urgent texts, the too-good deals, the "support" calls. One day a message nearly got him — and it wasn't carelessness, it was a good fake aimed at someone who grew up trusting people.

I didn't want him to stop using his phone. I wanted him to have someone to ask — someone who never gets tired of "is this real?", never makes him feel foolish, and always answers in plain English. So I built Ania: you send anything suspicious — a text, an email, a link, a photo — and it tells you what it sees and exactly what to do next.

"Not sure? Ask Ania." — the whole product in four words.

Every design choice comes from that: big buttons, short sentences, calm verdicts, and a promise with no asterisks — Ania will never ask for your password, a verification code, or payment. Anyone who does, even if they say they're Ania, is the scam.

Who we are

• Built and operated by: Nathaniel — design, code, and support all come from one desk.
• Contact: hello@getania.com — a real person reads every message.
• Running it for a group? The family organizer who set up your household manages its settings and members.

How Ania works — in plain English

Ania does four things: checks anything suspicious, helps with everyday tech trouble, walks you back from trouble if you already clicked, and can give your chosen family member a quiet heads-up about events. A signed-in household guardian can open the member's portal history view regardless of the privacy dial; the dial filters which event summaries appear, and the view omits submitted-message text and full Ania replies.

Where your words live

When you send Ania something, it is stored on Ania's hosted server. Each family can access only its own account and records; this is an online service, not a separate physical server for each family. Ania doesn't sell your messages or use them for advertising. Hosting, email, SMS, and—when enabled for your plan—AI providers may process information to provide their services, as explained in the privacy policy.

Your organizer can delete your member record from the active service store, and message history is cleared by an automatic 90-day retention sweep. Incident and alert screenshots are removed from active records when handled or resolved; rotating backups default to seven snapshots and a 30-day expiry threshold, both configurable. Expired copies are pruned at boot and daily while the service runs, even if nothing has changed; daily sweeps may leave an expired copy until the next sweep, and pruning waits while the service is stopped. Separate pre-migration recovery snapshots are not automatically pruned and may remain indefinitely. A signed-in household guardian can open the portal history view regardless of the privacy dial, but that view contains only dial-filtered event summaries—not submitted-message text or full Ania replies. See the privacy policy for details.

Where AI fits (and where it doesn't)

Ania checks messages in layers:

• The first layer is a rule engine built by hand — patterns of real scam tactics, checked on Ania's own server. It always runs, even with the internet's cleverest tricks in the message, and it works even if the AI layer is off.
• The second layer is an AI assistant — Claude (by Anthropic), with GPT (by OpenAI) as a backup. It reads screenshots, understands odd phrasing, and writes warmer replies. Which plans include it is on our plans page; even when it's on, your messages are sent only to analyze the one you sent — never to train models.
• Either way, the final gate is a set of hard safety rules the AI cannot override: never relay scam links or numbers, never give instructions that could help a scammer, always push you to official apps and websites you open yourself.

Your messages are never used to train AI models. They're sent only to analyze the message you sent, under those providers' applicable API terms. Hosting and delivery providers may also process data to provide Ania's service, as explained in the Privacy Policy.

What Ania isn't

Ania gives its best honest read — and says how confident it is. When it isn't sure, it says so and points you to the source that can be sure: your bank's official app, the number on the back of your card, the company's own website typed in by hand. It's a second set of eyes, not a replacement for them.

Questions, straight answers

• Can I see my history? Members can read their own full conversation history when signed in. A signed-in household guardian can open the member's portal history view regardless of the privacy dial, but sees only dial-filtered event summaries—not submitted-message text or full Ania replies. An append-only record of who has viewed that log is kept, and you can see who has viewed your own.
• What does my family see? A signed-in guardian can open the portal history view regardless of the privacy dial. The dial filters which event summaries appear; that view omits submitted-message text and full Ania replies. The portal keeps an append-only record of who has viewed each member's log; the family can see those records. Check the privacy policy before sharing sensitive information.
• What if Ania gets it wrong? Every reply shows its confidence, and the recovery guide ("🚨 I Messed Up") works from the first move — it never blames, it just helps you contain it fast.
• Who do I contact about privacy? Your family organizer first, or email hello@getania.com for anything about Ania itself.