Beckett

Beckett started as a personal Slack bot I built to make it easier to take and then actually benefit from notes. I found that tools like Notion often took as much time to organize as the notes took to write. I was also unimpressed with big-lab AI memory, which compresses away important details and keeps context trapped inside one product.

So I built Beckett’s predecessor, Atlas, as a conversational way to record thoughts and ideas, automatically categorize them, and tag relevant entities. Because it was bootstrapped, I kept the full-message context window small and loaded relevant information programmatically only when it was needed. When I mentioned a person or project, Beckett automatically surfaced related notes and past ideas.

As the product developed, I decided to make it multi-tenant, and brought that same context storage and retrieval system to other tools like calendar, tasks, trackers (habits, goals, and other trackable data), and projects. As an enthusiastic movie-watcher, avid reader, and amateur cook, I also included features for tracking films, shows, books, recipes, meal plans, and grocery lists. The guiding idea of Beckett is to create a single interface for your work, life, and home.

Today, Beckett is available on iPhone, web, desktop, and iPad, as well as through iMessage/SMS. I’m currently building MCP adapters for Claude and ChatGPT so those systems can use the same shared context.