What fai is

What is fai?

You open a session. You name it after what you're working on. You work with your AI as normal. When you're done, you seal the session.

That's it. That's fai.

Behind the scenes: everything that happened - your AI conversation, your file changes, your decisions - was captured. Next session, your AI reads everything from before. No re-explaining. No lost context.

What it is

CORE IDEA

What fai actually is

fai is your AI workbench - a portable, persistent layer that sits between you and every AI tool you use. It makes every AI you work with smarter, and stays yours regardless of which tools or models exist tomorrow.

Not a coding agent. Not a model router. Not a memory plugin. A workbench - owned by you, git-backed, yours to inspect, modify, or export anytime.

Three things that make it yours

PORTABLE

Travels with you

Your context travels with you - between projects, tools, and models. Start a session in VS Code, continue in Cursor. Switch from Claude to GPT. Your workbench comes along.

PERSISTENT

Session 50 > Session 1

Every session compounds on the last. Session 50 starts where session 49 ended. By month 2, your AI knows your naming conventions, your architectural decisions, your preferred patterns - not because you configured them, but because they emerged from how you work.

AI-AGNOSTIC

Works with any AI

Via MCP, any MCP-compatible AI reads your workbench. fai today supports Claude Code, Cursor, Windsurf, Cline, Codex, and more. New tools tomorrow: same workbench.

What grows when you use it

Your workbench accumulates three kinds of knowledge:

  • This session - what you're working on right now
  • This project - your stack, architecture decisions, naming conventions
  • You - your patterns, preferences, and workflow instincts across every project

None of this is configured. It appears as you work.

The scale journey

fai is where you start - your personal workbench (YOU^AI). As your team grows, it becomes a shared workshop (TEAM^AI). When you're shipping products, a studio (PRODUCT^AI). For large organizations, a platform (ENTERPRISE^AI).

Same architecture underneath. Vocabulary adapts to where you are.

Next steps

By curiosity level:

On this page