Alabasta is my bet on what comes after Linear and Notion: a workspace where AI is infrastructure, not a feature. The goal isn’t a chatbot in a sidebar — it’s a system that proposes the next action, surfaces context, and keeps a living memory of how a product actually evolved.
The thesis
Teams shred their product intelligence across Jira, Linear, Slack, Notion, and Figma. Decisions, tradeoffs, and blockers vanish into chat history. Alabasta aims to be the operating system for product reasoning — a real-time knowledge base of:
- decisions and the tradeoffs behind them
- threads, blockers, and ownership
- issues, roadmaps, and implementation state
What makes it different
- Predictive, not reactive — the system proposes metadata, related work, and risk inline as you type, instead of waiting to be asked.
- Command-first UX — everything is keyboard-accessible, fuzzy-searchable, and chainable. Low visual noise, tight typography, motion discipline.
- Context everywhere — hover any entity (issue, member, release, metric) and pull its context without leaving the page.
Where it stands
In active development. The hard parts — a fast, optimistic data layer on Convex, a TipTap-based block editor with slash commands and image paste, and the AI orchestration layer — are the ones I’m building first, because they’re what the product lives or dies on.