Most people use AI the same way: they open a chat window, ask a question, get an answer, and close the window. The next time they need something, they start from scratch.
That works fine for one-off tasks. It doesn’t work for anything that requires continuity — building up knowledge over time, maintaining a consistent understanding of a project, or thinking through a complex problem across multiple sessions.
The system I use every day is built around two tools: Obsidian and Claude. Together, they function as a persistent, AI-augmented thinking environment. Here’s how it works — and why it’s the kind of thing that can be built for a business, not just a developer.
The Problem: AI Has No Memory
Every time you start a new conversation with an AI, it knows nothing about you, your work, your projects, or your preferences. The context you have to rebuild every session is friction — and friction accumulates.
The answer isn’t a better AI. It’s a better system around the AI.
The Setup
Obsidian is a markdown-based note-taking application that stores everything as plain text files on your machine. There’s no proprietary format, no cloud lock-in, no subscription required to access your own notes. It’s designed for building a connected knowledge base — a “second brain” — where ideas link to each other and accumulate over time.
Claude operates as the intelligence layer on top of that knowledge base. It reads the wiki, synthesizes information across documents, answers questions with citations, writes new pages, and updates existing ones as understanding evolves.
The result: a knowledge system that grows smarter over time because the AI is maintaining it, not just answering questions from a blank slate.
What It Looks Like in Practice
The wiki contains everything I need to carry forward: project notes, design decisions, operational runbooks, research threads, open questions. Each topic has its own page. Related pages link to each other.
When I start a session, Claude reads the current state of the wiki — the live session file, the recent log, the index — and knows exactly where things stand. When I ask a question, it searches the wiki for relevant context before answering. When the answer is valuable, it gets filed as a page, not lost in a chat transcript.
The knowledge compounds. A decision made three months ago is documented and searchable. A pattern observed across multiple projects gets its own wiki page. A research thread that started as a question evolves into a reference document.
Why This Matters for a Business
The same architecture applies anywhere people need to think with continuity — any team that generates knowledge, makes decisions, and needs those decisions to be accessible later.
A few examples of where this pattern is useful in a business context:
Strategic planning and decision logging. Every significant decision your organization makes involves context that disappears into email threads and meeting notes. A structured knowledge base with an AI layer makes that context retrievable and useful.
Project operations. Complex projects accumulate decisions, open questions, and institutional knowledge. Building that into a maintained wiki instead of letting it scatter means the next person to join the project — or the next engagement with a similar client — starts from a much stronger position.
Operational documentation that stays current. Most operations documentation is wrong within six months of being written. An AI-maintained wiki gets updated as procedures change — not as a separate documentation project, but as a byproduct of doing the work.
Building a custom AI-augmented workflow system is one of the four things we do at NewThink.ai. The Obsidian + Claude pattern is one example — the right setup for your organization depends on what you’re trying to accomplish. Let’s find out what that is.