# Discovery: The Second Brain That Runs Our Studio

> Our studio runs on a plain-text knowledge base in Obsidian connected to Claude. Clients, tasks, and decisions in one place, with automated daily reports. Here's what we learned about building a second brain that actually sticks.

Source: https://pulsarinteractive.ai/blog/second-brain-obsidian-claude
Published: 2026-10-15 by Pulsar Interactive, Pulsar Interactive
Updated: 2026-10-15

Every small team has the same hidden problem: the business lives in too many places. Client notes in one app, tasks in another, decisions in chat threads, and the real context in someone's head.

We hit this ourselves. So we built a "second brain" for our own studio, and connected it to AI.

> **[ MEDIA SLOT 1 · IMAGE ]** The home dashboard of a second brain (made-up content only, never a live vault). Replace this block with the image.

## What we found

AI assistants are only as useful as what they can see. Asking Claude "what's the status of this client?" is pointless if the answer is spread across five apps it can't reach.

We also found the opposite risk: once you connect AI to your notes, you have to decide what it should *never* see.

## What we built

Our studio runs on a plain-text knowledge base in Obsidian, kept in version control:

- **One thin index** points to everything else: clients, tasks, decisions, finances, travel.
- **One status file per client**, structured so both people and AI can read it.
- **Daily logs**, many of them written automatically by a scheduled report that reviews the whole vault every morning and flags anything that needs attention.
- **Custom AI skills** for the jobs we repeat: a status check, quick capture of ideas, a daily summary.
- **A connection to Claude**, so we can ask questions in plain language and get answers grounded in our own files.

> **[ MEDIA SLOT 2 · IMAGE or VIDEO ]** Asking Claude a question and getting an answer from the vault (made-up content). Replace this block, or delete it.

## What we learned

**1. The tools take a day. The habits take months.** Setting up the vault and connecting AI is quick. Getting into the rhythm of capturing things, reviewing them and trusting the system is the real work. That's the part worth teaching.

**2. Decide what can't go near AI, first.** Before connecting anything, every person should answer one question: *what must never go into this?* For a clinician it's patient data; for a business it may be bank details or HR files. Those boundaries shape the whole setup.

**3. Plain text lasts.** Notes in simple files outlive any app. If a tool disappears, your knowledge doesn't.

**4. A daily report changes behaviour.** A short summary waiting every morning does more for follow-up than any reminder app.

## Who this is for

- **Founders and small teams** whose business lives in too many apps and too many heads.
- **Consultants and advisors** who need context on many clients at once.
- **Professionals** who want AI help without exposing sensitive information.

## We can set this up for you and your team

We set up a second brain that fits how you already work, and teach the habits that make it stick:

- **Session one:** we build your vault together, connect it to Claude, sync it to your phone and set up your dashboard.
- **A week or two on your own,** using it day to day.
- **Session two:** we fix what didn't fit and add the skills and reports you actually need.

We start with your boundaries: what goes in, and what never does. If your team's knowledge is scattered across apps and inboxes, **[tell us how you work today](https://pulsarinteractive.ai/#contact)**.