Technology & Trust
How Aging Sidekick actually works — where your family’s information lives, how the AI is tested before it talks to you, why every answer cites a source you can check, and what a purpose-built care assistant does that a general chatbot cannot.
Aging Sidekick helps family caregivers get organized and find their next step — at their own pace.
Why not just use ChatGPT? What a purpose-built care assistant does differently
General chatbots are impressive, but they do not know your mother. Here is what a care assistant built for family caregivers does differently — real context about your person, cited answers from vetted sources, tested safety behavior, and a record that builds over time.
Where your family's information actually goes (and where it never goes)
You are trusting us with details about your parent's health, home, and daily life. Here is exactly how that information is stored, what the AI sees when you ask a question, and why it is never sold and never used to train AI models.
How we test the AI before it talks to your family
Before an answer ever reaches a caregiver, our assistant has to pass a 118-scenario evaluation suite — everyday questions, messy records, safety cases, and emergencies that must reach a human. Here is how the testing works and what the numbers say.
Why the assistant only cites sources you could look up yourself
AI chatbots can invent agencies, phone numbers, and rules that sound right. Ours cannot cite the open web at all — its research runs against 27 vetted sources like Medicare.gov, the National Institute on Aging, and the Eldercare Locator, and every recommendation carries a link you can check.
The safest thing an AI can say is 'call your doctor'
Our assistant is built to decline medical diagnoses, route emergencies to 911 and the 988 crisis line, ask before guessing, and admit what it does not know. Here are the four guardrails — and how we verify they hold before every release.