Locul

Locul turns the notes, docs, and dictations you already create into a memory your AI can use. It’s for people who want their AI to answer with what’s true today, without manually organizing, tagging, or keeping a “second brain” refreshed.

Locul is built around a simple idea. Your inputs become a memory store, and Locul keeps that memory up to date as things change, so you don’t get stale answers from last week’s version.

• Local-first memory. Your memory lives on your machine, with an opt-in cloud AI mode that uses your own key.
• Reads what you already have. It pulls from files (Markdown and PDF), Notion, an Obsidian vault, and dictations you do through Wispr Flow (and Contextli).
• Auto-updates changed content. On a regular interval, it re-reads inputs and updates memories that are out of date while preserving what changed.
• Built for AI workflows. It wires into the AI tools you already use (Claude, ChatGPT, Cursor, Codex, OpenClaw) so they can access the current memory when they need details.
• Keeps your public voice consistent. It can pull your own LinkedIn posts through your account, then store them locally (back catalogue first, then refreshes).

What sets it apart is that you don’t “feed” it continuously, and you don’t have to maintain folders or tags. You connect it once, then it handles ongoing indexing and updates. It’s also designed for people who already use AI in their day-to-day work and want their AI to stop forgetting product decisions, pricing, positioning, and other changing details.