We started building Guidy because we kept running into the same frustrating experience with AI: It can be incredibly smart in the moment, but still feel like it barely knows you over time. You explain your projects, preferences, goals, important people, and decisions — and eventually some of that context gets lost, becomes outdated, or has to be explained again. That made us ask a bigger question: What would AI look like if memory were a core capability, not just an add-on? At first, we thought the problem was simply “AI needs to remember more.” But while building Guidy, our thinking changed. We realized useful long-term memory isn’t about storing everything forever. A Personal AI needs to know: what is worth remembering how different memories relate to each other when information has changed what should be updated or forgotten when past context is actually relevant and when it should proactively help without becoming intrusive That led us to build MemG, our AI-native long-term memory engine, and then build Guidy on top of it. The product also evolved from being mostly a memory-aware assistant into something more proactive. Instead of only asking: “How can AI answer this better?” we started asking: “Given what AI already knows about you, what should it help you with next?” That shift — from answering, to remembering, understanding, and acting — became the foundation of Guidy. We’re still early, and that’s exactly why we’re launching now. We want real users to help us answer the questions we can’t solve in a lab: What should a Personal AI remember about you? What should it forget? And what would make proactive AI genuinely useful in your daily life? Would love to hear how you think about this. From Memory to Action.

Forum Threads

p/guidy-2

•

18d ago

When should a proactive AI act — and when should it stay quiet?

We ve been thinking about this a lot while building Guidy.

Long-term memory can help an AI understand your ongoing projects, past decisions, preferences, and unfinished tasks. But memory alone doesn t necessarily make an AI useful.

The harder question is: when should it actually do something with that memory?

For example, imagine an AI remembers that:

View all

Guidy

Personal AI with long-term memory that acts for you.

Personal AI with long-term memory that acts for you.

Forum Threads

When should a proactive AI act — and when should it stay quiet?

Forum Threads

When should a proactive AI act — and when should it stay quiet?