Anchor is personalized AI support for older adults and their families. Grace uses routines, memories, interests, relationships and everyday context to make conversations more familiar and useful, then turns what was shared into simple caregiver reports and trends over time. We’re starting with families navigating memory changes, Alzheimer’s and dementia, with a broader goal of supporting independence and aging in place.

How did Astra change the scope or ambition of what you built?

Astra changed Anchor from an early concept and technical proof of concept into something I could actually put in front of people. Using Astra through Codex, I was able to expand the scope into personalized conversation flows, memory and context handling, caregiver-report concepts, safety guardrails, the surrounding UX, and a public pre-alpha demo. The biggest shift was moving from “Can this technically work?” to “What would this feel like as a real product?” In less than a week, using only the Plus plan, Astra helped me turn Anchor into a tangible release with a live demo and waitlist. That speed let me explore the full two-sided experience for both the older adult and the caregiver, not just the AI interaction itself.

Hey Product Hunt 👋 I’m Tyler, the maker of Anchor. I started thinking about Anchor after seeing loved ones experience Alzheimer’s and dementia. What stayed with me wasn’t only the memory loss. It was everything around it: loneliness, changes in routine, and how difficult it can be for families to know how someone is doing between visits. Even when a family cares deeply, there are still a lot of hours when they simply can’t be there. That’s the problem I started exploring with Anchor. Grace, the AI behind Anchor, is designed to use context like routines, memories, interests, relationships and everyday life to make conversations feel more familiar and relevant. The other side of Anchor is for the family. What someone shares can become simple caregiver context around things like their day, mood, routine and anything that may be worth following up on. One conversation. Two experiences. For the older adult, a personalized interaction built around their life. For their family, a little more context about the moments they can’t always see. What I’m launching today is still a pre-alpha proof of concept. The demo lets you simulate a conversation with Grace, see how personal context can shape the interaction, and then view an example caregiver report based on what was shared. The longer-term vision is broader than dementia. I’m interested in how Anchor could help older adults stay connected, supported and independent as they age, while helping families remain involved without taking that independence away. For the GPT-6 Astra Challenge, I used GPT-6 Astra through Codex to explore and build this experience, particularly the conversation flow, personalization, caregiver-report concept and safeguards around what the AI should and shouldn’t infer. Astra helped me take Anchor from an early concept and technical proof of concept to a tangible pre-alpha release in less than a week, using only the Plus plan. That speed let me move beyond architecture and experimentation into something people can actually try, react to and help shape. I’m also opening the Anchor waitlist with this launch. If you care for an aging parent, grandparent or loved one, I’d especially love feedback on two things: * Would an experience like this be useful between your own calls and visits? * Does the caregiver-report concept provide useful context while still respecting the older adult’s dignity and independence? Thanks for checking out Anchor. It’s early, but it’s a problem I care deeply about getting right.

How did Astra change the scope or ambition of what you built?

Anchor

Personalized AI for aging in place and caregiver insight

Personalized AI for aging in place and caregiver insight