One place for all your AI conversations.
After years of building AI solutions for companies, I've decided to start building products in public. My first project is an AI Memory Vault a workspace that helps developers keep all their AI conversations in one place instead of losing them across ChatGPT, Claude, Codex, and other tools.
The goal is simple: make AI work easier to search, organize, and continue.
I'll be sharing progress, technical decisions, challenges, and lessons as I build. Looking forward to learning from the Indie Hackers community.
The part about continuing conversations across ChatGPT Claude and Codex is what grabbed me. That's a problem I keep seeing get worse as people bounce between tools.
I can imagine a developer asking ChatGPT something like "how do people keep track of prompts across different AI tools" and stumbling onto this without ever hearing the name before. If that first look makes it feel like a place other builders already rely on they'll be a lot more willing to move years of conversations into it.
Curious if you've talked to people who are already juggling multiple AI tools every day or if you're still testing whether this pain is common enough.
Love this direction — it solves a very real pain.
The “lost context across tools” problem is only getting worse as people use multiple AI systems daily. Having a single place to search, organize, and continue conversations feels like a natural next step, especially for developers juggling ChatGPT, Claude, Codex, etc.
Building in public is also a smart move here — you’ll get fast feedback on what features actually matter (search, tagging, context linking, maybe even auto-summaries).
Curious how you’re thinking about:
cross-platform ingestion (APIs vs manual import)
structuring memory (threads vs semantic chunks)
and privacy, since convo data can be sensitive
Definitely a project worth following 🚀
The “temporary chat but I still want to keep this” tension is very real. I do the same dance of not wanting providers to train on everything, then immediately regretting it when there’s a snippet or mini‑framework I wish I’d saved.
Treating these chats as reusable knowledge instead of disposable Q&A feels like the right mental model. Curious how far you’re planning to go beyond storage/search — e.g. recalling past threads by task, tagging conversations across tools, or helping people reconstruct a chain of reasoning weeks later. That’s where an actual “vault” becomes way more valuable than just another notes app.
I like that you're treating AI conversations as knowledge instead of disposable chats.
The interesting challenge is making the vault more than searchable history. The real value is helping people pick up complex work weeks later without having to reconstruct all the context from scratch.