Where your family's information actually goes (and where it never goes)

To be useful, Aging Sidekick asks you to write down sensitive things: your father's memory lapses, the fall he did not tell anyone about, what the family can afford. You should know exactly what happens to that information — in plain language, not a privacy policy's.

Where your information lives

Everything you enter — profiles, assessments, notes, conversations — is stored in a database where every single record is tagged to your account and walled off by row-level security. Think of a filing cabinet where every folder has its own lock, and your key opens only your folders. This is not a policy we ask employees to follow; it is enforced by the database itself on every request. Documents and audio you upload are stored the same way, in storage paths scoped to your account.

Your data is encrypted in transit and at rest, access-controlled, never sold. That last part is worth saying twice: our business is subscriptions, not data. There is no advertising side of the house, no data-broker relationship, no "partners" clause that quietly means selling lists.

What the AI sees when you ask a question

When you ask the assistant something, it is sent the care profile you built — the assessments, concerns, and plans for the person you care for — plus your conversation. That context is what makes the answer specific instead of generic. What it is not sent: your email address or your account identifiers. The AI gets the case file, not your identity.

Why we ask for so much detail

A fair follow-up: if this information is so sensitive, why collect it at all? Because it is the difference between advice and help. "Talk to her doctor about fall risk" is advice. "Her bathroom is upstairs, she is already using a walker, and her last assessment flagged nighttime trips — here are the three questions to ask her doctor this week" is help. Every field we ask for exists to make an answer more specific, and you decide how much to share. Start with a name and one concern; add detail when the answers earn it.

Your conversations never train the AI

This is the fear underneath most questions about AI and privacy: "If I tell it about my mom, am I feeding some tech company’s next model?" No. Aging Sidekick runs Claude through Amazon Bedrock, an enterprise AI platform built for exactly this concern. Under Bedrock’s data commitments, your prompts and the assistant’s responses are not shared with the company that makes the model and are not used to train or improve it. And Aging Sidekick itself never uses your family’s information to train AI models either.

What we do not claim

Honesty is part of trust, so here is the fine print in normal type. We are a consumer service, and we are not a HIPAA-covered entity — see our Consumer Health Data Privacy Notice for how we handle health-related information, and our Privacy Policy for the complete picture. If a company selling software tells you privacy is simply "handled," ask them the questions this post just answered.

Deleting your information

Your records exist for one reason: so the assistant can help you. When a profile is no longer needed, you can delete it, and the chat history and notes attached to it are deleted with it. Leaving entirely works the same way — this is your family’s working record, not ours.

• Encrypted in transit and at restStandard strong encryption protects your data moving and stored.
• Isolated by accountRow-level security walls every record off from every other account.
• Never soldNo advertising, no data brokers, no list-sharing. Subscriptions are the business.
• Never used to train AINot by us, and not by the model provider — enforced by Amazon Bedrock’s data commitments.
• DeletableDelete a profile and its chats and notes go with it.

Curious how the assistant behaves with all that context — and how we verify it? Start with how we test the AI before it talks to your family or the overview in why not just use ChatGPT.