@vosRemember This: semantic search for your own notes, from a MongoDB engineer
Hi all, I'm Pauline, a software engineer in Amsterdam. By day I'm a senior engineer at MongoDB, where I maintain open-source software for the PHP ecosystem. On the side I'm building Remember This, an app for saving notes, photos and voice notes and finding them later by describing what you remember, instead of hunting for the right keyword or folder.
It's in closed beta right now. Under the hood it uses MongoDB Atlas Vector Search with Gemini and VoyageAI, and I build most of it with Amp, Claude Code and Codex. I'm still figuring out how to get the most from each of them, and what a good search experience feels like when people only half remember things.
What I'd love from this community: feedback from anyone who has shipped SaaS to non-technical users and spread out to enterprise and technical users over time.
Comments3
My brain needs this app
Welcome to Lounge, Pauline. Searching by half-remembered descriptions across text, photos and voice notes is a hard retrieval problem, since each needs its own embedding path. Are you embedding images and audio directly with Voyage's multimodal model, or captioning and transcribing first, and how does that change recall?
Love the idea. It might be a great product, good luck!
3 comments
My brain needs this app
Welcome to Lounge, Pauline. Searching by half-remembered descriptions across text, photos and voice notes is a hard retrieval problem, since each needs its own embedding path. Are you embedding images and audio directly with Voyage's multimodal model, or captioning and transcribing first, and how does that change recall?
Love the idea. It might be a great product, good luck!