Building Your AI-Powered Second Brain: How Claude Code + Obsidian Changed Everything About My Knowledge Management

After 10 years of scattered notes, forgotten ideas, and lost insights across dozens of apps and platforms, I discovered something that fundamentally changed how I think, learn, and work.

After 10 years of scattered notes, forgotten ideas, and lost insights across dozens of apps and platforms, I discovered something that fundamentally changed how I think, learn, and work.

I’m Alek, a Product Manager at C&C (an Apple Premium Partner across 8 European countries), where I build internal tools for retail, education, and B2B teams. I work daily with Vue, Nuxt, Supabase, and n8n, mostly using AI-assisted coding through Cursor and now Claude Code.

This isn’t academic theory — it’s based on real projects with real deadlines, managing development teams across multiple countries, and building enterprise software that serves thousands of users. You can find more about my work and projects at alekdob.com

What You’ll Learn Today

• How to transform scattered thoughts into an interconnected knowledge system
  • Why Obsidian + Claude Code creates the perfect AI-powered second brain
  • My proven organizational system using tags and linking strategies
  • How AI can become your thinking partner beyond just coding
  • Practical setup steps to build your own self-feeding knowledge ecosystem
  • Real examples from managing international software projects

The Knowledge Management Crisis We All Face

Here’s a sobering statistic: the global knowledge management market was valued at $773.6 billion in 2024 and is projected to reach $2.1 trillion by 2030. Yet 75% of knowledge workers report feeling overwhelmed by information, and the average professional spends 2.5 hours daily just searching for information they know exists somewhere.

I lived this reality. Notes scattered across Apple Notes, Notion, Google Docs, Slack messages, and email threads. Ideas captured on napkins, in voice memos, or worse — lost entirely because I didn’t have the right app open at the right moment. The cognitive overhead of managing this chaos was eating into my actual thinking time.

Then I discovered the combination that changed everything: Obsidian as the foundation, enhanced by Claude Code as the AI thinking partner.

💡 Key Concept: Your brain is for having ideas, not storing them. A second brain should enhance your thinking, not compete with it.

Use this when you need to distinguish between active thinking and knowledge storage

Why Obsidian is the Perfect Foundation

After testing dozens of knowledge management tools — from Roam Research to Notion, Tana, Reflect, Mem.ai, Constella, Capacities, SuperNotes, Kinopio, Heptabase, Scrintal, LogSeq to Remnote — Obsidian emerged as the clear winner for four critical reasons:

1. Local-First Philosophy: Complete Data Ownership

Your notes live on your device in simple Markdown files. No vendor lock-in, no subscription anxiety, no “what happens if the company shuts down” concerns. I can open any note with any text editor, now or 20 years from now. This isn’t just philosophical — it’s practical. When you’re building enterprise software across 8 countries, data sovereignty matters.

2. Markdown Format: AI-Friendly and Future-Proof

Every major AI system understands Markdown perfectly. When Claude Code reads my notes, it sees clean, structured content without proprietary formatting noise. This creates a seamless bridge between human-readable content and AI processing.

3. Cross-Platform Excellence That Actually Works

The iOS, Android, and desktop apps are blazingly fast and feature-complete. I can capture thoughts on my iPhone while traveling between C&C offices in Stockholm, edit on my iPad while waiting for flights, and do deep work on my MacBook in Puglia. The sync is instant and reliable.

The Mobile Revolution Nobody Talks About

Here’s what most PKM tools get wrong: mobile isn’t an afterthought — it’s where 60% of ideas are born. I’ve tested everything from Notion to Tana, and while Tana is powerful on desktop, its mobile experience is sluggish and frustrating. Obsidian? It’s a different story entirely.

The mobile app is a genuine masterpiece of engineering. Real-time sync that actually works in milliseconds. Full feature parity with desktop. Native performance that feels like a first-party Apple app. When I’m in a meeting with partners in Milan and need to quickly capture an insight, I don’t want to wrestle with a laggy interface. I need something that gets out of my way and lets me think.

This isn’t just convenience — it’s critical for knowledge capture. Ideas don’t wait for you to be at your desk. They come during commutes, in coffee shops, between meetings. A second brain that only works at your desk is only half a brain.

4. Simple Yet Infinitely Scalable

Beginners can start with basic note-taking, while power users can build complex automated workflows. I started with simple project notes and now have a knowledge graph with thousands of interconnected concepts.

💡 Key Concept: The best knowledge management system grows with you rather than forcing you to adapt to it.

Use this when evaluating any long-term tool or system

My Organizational System: Simple but Powerful

After experimenting with complex folder hierarchies and elaborate naming schemes, I discovered that simplicity wins. My entire system runs on seven core tags:

The Tag Taxonomy That Actually Works

• #human: People I work with, meet, or learn from

• #project: Work initiatives, side projects, business ventures

• #bookmark: Resources, articles, tools worth revisiting

• #wiki: Reference material and documentation

• #goal: Aspirations and objectives at any timeframe

• #todo: Actionable items and tasks

• #atomic-note: Single-concept insights and learnings

• #meeting: Meeting notes and discussions

• #area: Core life areas and fundamental concepts

  

The Simplicity Principle

Here’s the trap most people fall into: they create elaborate property systems with 50 fields and complex hierarchies. Then they open the app on their phone, see all those required fields, and… close it again. The friction kills the habit.

My rule: if I can’t add a note in under 10 seconds on mobile, the system is too complex. Tags are instant, intuitive, and flexible. No mandatory fields, no complex forms, just capture the thought and add context with a simple #tag.

This simplicity is what makes the system sustainable. After 18 months, I still use it daily because it doesn’t fight me. It adapts to how I think, not the other way around. Remember: the best system is the one you actually use, not the one with the most features.

This creates natural hierarchies without rigid structures. My “Alek Dobrohotov” note (#human) connects to “alekdob.com” (#project), which links to specific goals (#goal) and actionable todos (#todo). The system breathes and evolves organically.

Double Bracket Linking: Where Connections Come Alive

Obsidian’s double bracket linking is where the magic happens. Instead of filing notes in folders, I create relationships between concepts. When I write about implementing OAuth in our Flow ERP, that note naturally links to:

• Supabase Authentication (technical implementation)

• C&C Security Requirements (business context)

• User Experience Design (product considerations)

• International Compliance (regulatory aspects)

These connections create a web of knowledge that mirrors how we actually think — through associations and relationships, not rigid categories.

💡 Key Concept: Knowledge isn’t hierarchical — it’s networked. Your external brain should reflect this reality.

Use this when designing any information architecture

The Knowledge Graph: Watching Your Brain Grow

One of Obsidian’s most compelling features is the visual knowledge graph. It shows your notes as nodes connected by links, creating a living map of your thinking. Over 18 months, I’ve watched mine evolve from scattered islands into interconnected continents of knowledge.

The graph reveals patterns I never consciously created. My notes about AI development cluster around productivity tools, which connect to project management insights, which link back to team communication strategies. The visual representation often surfaces unexpected connections — like how my notes on Italian business culture inform my approach to international software deployment.

When I’m stuck on a problem, I often browse the graph view around related concepts. It’s like having a conversation with my past self, discovering insights I’d forgotten and connections I hadn’t fully explored.

The Claude Code Integration: Where Magic Happens

This is where my second brain becomes truly intelligent. Claude Code doesn’t just see my code — it understands my entire context. Here’s how I set it up:

Technical Setup (5 Minutes)

1. Open your Obsidian vault folder with your preferred IDE or terminal

2. Run claude from terminal to initialize Claude Code in your project

3. Run init to create the CLAUDE.md file where you can:

• Add crucial information about yourself

• Link to important files that define who you are

• Explain your tag system and what each tag means

• Describe your folder structure

• Note that you use Zettelkasten methodology

• Provide any context that helps Claude Code navigate and search effectively

The CLAUDE.md file becomes your AI’s instruction manual. You can write whatever you want — this is where you teach Claude Code how to understand your specific system.

The Power of Custom Subagents

Here’s the advanced technique: create specialized subagents that work with your second brain. I’ve built:

• A Medium article writer that understands my writing style and audience

• A Substack growth assistant that analyzes engagement patterns

• A project documentation agent that maintains consistency across all my notes

• A learning synthesis agent that connects new information to existing knowledge

These agents don’t just access your knowledge — they become extensions of your thinking process, each specialized for specific types of work. (I’ll cover this in detail in my next article.)

That’s it. Claude Code now has access to your complete knowledge base, understanding your projects, goals, relationships, and thinking patterns.

What Claude Code Knows About Me

When I ask Claude Code to help with Flow ERP development, it already knows:

• My role as Product Manager at C&C

• Our tech stack (Vue, Nuxt, Supabase, n8n)

• Current project challenges and goals

• Team structure across 8 European countries

• Previous decisions and their reasoning

• My Life goals and passions

• My career ambitions

This context transforms every interaction. Instead of explaining background information repeatedly, we dive straight into solving specific problems with full awareness of constraints and objectives.

The Global Agent Revolution

Here’s where I’ve taken this to the next level: I’ve created a custom global agent specifically designed to investigate my second brain. This isn’t just a local tool — it’s available everywhere in Claude Code, across all my projects.

What makes this incredibly powerful is Claude Code’s CLI nature. It can work across multiple folders simultaneously. So when I’m deep in the Flow ERP codebase, my agent can:

• Pull context from my Obsidian vault

• Reference meeting notes from last week

• Find similar problems I’ve solved before

• Connect technical decisions to business objectives

Imagine this scenario: I’m implementing a new payment system. My global agent can simultaneously:

1. Check the project folder for existing payment code

2. Search my second brain for payment-related decisions

3. Find meeting notes where stakeholders discussed requirements

4. Identify team members with relevant expertise

5. Surface compliance notes from different countries

This creates perfect context. The AI doesn’t just see code — it sees the complete picture of why, who, when, and how.

Building Your AI Team

The real magic happens when you create specialized agents. My global agent for second brain investigation is just the beginning. You can build:

• Content creation agents that know your voice

• Research agents that understand your interests

• Project management agents that track your methodologies

• Learning agents that reinforce your study patterns

Each agent becomes a specialist that deeply understands one aspect of your work while having access to your entire knowledge context.

The Future Investment

As AI context windows expand, this approach becomes even more powerful. Every note I add today is an investment in tomorrow’s AI capabilities. We’re not just building a knowledge base — we’re training our personal AI assistant with years of contextual understanding.

The Self-Feeding Ecosystem

Here’s where it gets really interesting. As Claude Code helps me work, it can also help me document and connect that work back to my knowledge base. After implementing a new authentication flow, I might ask:

“Based on what we just built, create a note about OAuth implementation patterns that connects to my existing notes on security and user experience.”

Claude Code writes the note, suggests connections to existing content, and even prompts me with questions: “Did this implementation reveal any insights about international user behavior patterns you’ve been tracking?”

💡 Key Concept: The best AI assistants don’t just execute tasks — they help you think better about the work itself.

Use this when evaluating AI tools for productivity

Beyond Code: Programming Your Life

While Claude Code excels at development tasks, its access to my complete knowledge base makes it valuable for every aspect of my work:

Strategic Planning Sessions

“Looking at my project goals and recent meeting notes, what patterns do you see in team feedback about our current development process?”

Learning Integration

“I just read this article about microservices architecture. Based on my notes about Flow ERP’s current structure, how might these concepts apply to our next iteration?”

Decision Documentation

“Help me create a decision log entry about choosing Vue over React for our new admin interface, connecting it to our existing technical choices and team expertise notes.”

Weekly Reviews

“Review my journal entries from this week and suggest connections to my longer-term goals. What should I be thinking about for next week?”

The AI becomes a thinking partner that remembers everything and helps surface insights across the entire knowledge graph.

Practical Use Cases: Real Examples from the Field

Project Documentation That Writes Itself

When planning Flow ERP features, I capture initial thoughts in atomic notes tagged #project and #goal. As development progresses, these connect to:

• Technical implementation notes

• User feedback observations

• Performance metrics

• Team retrospective insights

Claude Code helps weave these scattered observations into coherent project documentation, identifying gaps and suggesting areas for deeper exploration.

Meeting Notes That Connect to Action

After international team meetings, I create notes tagged #human for key participants and #todo for action items. The AI helps me:

• Connect discussion points to existing project contexts

• Identify recurring themes across multiple meetings

• Suggest follow-up questions based on previous conversations

• Link decisions to relevant technical or business constraints

Learning Journals That Surface Patterns

I maintain ongoing notes about productivity techniques, development insights, and management approaches. Claude Code helps identify patterns: “Your notes suggest you’re most productive with Vue development in morning sessions, but struggle with complex Supabase queries after 3 PM. Have you considered time-blocking your technical architecture work?”

💡 Key Concept: The compound value of documented experience only emerges when connections between experiences become visible.

Use this when building any learning or improvement system

The Compound Effect: Why This System Multiplies

Knowledge management isn’t linear — it’s exponential. Each note doesn’t just add to your collection; it potentially connects to every other note, creating network effects that multiply the value of your entire system.

Better Decision Making

When facing a new challenge, I can instantly access:

• Similar situations I’ve handled before

• Outcomes of different approaches

• Team member expertise and availability

• Technical constraints and opportunities

• Business context and priorities

This comprehensive context leads to faster, more informed decisions.

Accelerated Learning

New information doesn’t exist in isolation. When I learn about a new development technique, my system immediately shows me:

• Where it might apply in current projects

• Similar patterns I’ve used successfully

• Team members who might be interested

• Potential integration points with existing tools

Enhanced Creativity

The unexpected connections surfaced by the knowledge graph often spark creative solutions. A note about Italian design principles connects to user interface insights, which links to accessibility considerations, revealing a new approach to our international user experience.

Reduced Context Switching

Instead of reconstructing context every time I work on a project, the full background is immediately available. This eliminates the cognitive overhead of “getting back up to speed” and lets me stay in deep work longer.

Getting Started: Your Action Plan

Ready to build your own AI-powered second brain? Here’s your step-by-step roadmap:

Week 1: Foundation Setup

1. Download Obsidian from obsidian.md

2. Create your vault in a location you’ll remember

3. Install essential community plugins:

• Calendar (for daily notes)

• Smart Connections (AI-powered linking suggestions)

• Templater (for consistent note structures)

4. Create your first note: “Your Name.md” with basic information about yourself

Week 2: Establish Patterns

1. Create templates for common note types:

• Meeting notes

• Project overviews

• Learning journals

• Daily reflections

2. Start using the core tags: #human, #project, #bookmark, #wiki, #goal, #todo, #atomic-note, ….

3. Practice double-bracket linking — connect every new note to at least two existing notes

Week 3: AI Integration

1. Set up Git in your Obsidian vault folder

2. Connect Claude Code to your repository

3. Start asking AI to help with note connections and insights

4. Experiment with AI-assisted content creation and organization

Week 4: Optimization and Habits

1. Review your knowledge graph and identify strong/weak connection areas

2. Establish daily and weekly review routines

3. Create automated templates for recurring note types

4. Begin using AI for strategic thinking and pattern recognition

Essential Starting Principles

Start small: Begin with simple notes about your current projects. Don’t try to backfill years of information immediately.

Link liberally: When in doubt, create a connection. Over-linking is rarely a problem; under-linking always is.

Think in concepts: Each note should contain one main idea that could be useful in multiple contexts.

Trust the process: The real value emerges after you have 50+ connected notes. Be patient with the initial setup period.

💡 Key Concept: Systems thinking beats goal-focused thinking for long-term knowledge building.

Use this when building any sustainable habit or practice

Mobile-First Habits

Quick Capture Protocol: When on mobile, I follow this pattern:

1. Open Obsidian (instant, thanks to its speed)

2. Create note with descriptive title

3. Add one relevant tag

4. Write the core idea (even if it’s just 2 sentences)

5. Save and move on

Later, during desktop sessions, I enhance these notes with connections and expand on ideas. But the critical capture happens in the moment, friction-free.

The Future of Thinking

We’re witnessing the emergence of human-AI collaborative intelligence. The combination of a well-structured external brain with AI that understands your complete context creates possibilities that neither humans nor AI could achieve alone.

My Obsidian vault isn’t just storing information — it’s actively participating in my thinking process. Claude Code isn’t just helping me code — it’s helping me connect ideas, spot patterns, and ask better questions.

This is personal knowledge management evolved for the AI age. It’s not about replacing human thinking, but about creating the conditions where human creativity, AI processing power, and structured information can work together seamlessly.

The global knowledge management market is growing because information overwhelm is getting worse, not better. But with the right system and the right AI partnership, that same explosion of information becomes an advantage rather than a burden.

What You’ve Learned

You now know how to:
• Set up a future-proof knowledge management foundation with Obsidian
• Create an organizational system that scales with your thinking
• Integrate AI assistance that understands your complete context
• Build a self-reinforcing cycle of documentation and insight
• Transform scattered information into networked knowledge

The next time you capture an idea, document a decision, or learn something new, you’ll be ready to make it part of a system that grows more valuable with every addition.

Your second brain awaits. The only question is: what will you teach it first?

Found this helpful? There’s more where this came from. Check out my other articles on building enterprise tools and AI-assisted development.

👉 Follow me on Medium and Substack for weekly insights on Product Management and modern development workflows.

💬 Drop a comment below — I read and respond to every one. What’s your biggest challenge with knowledge management? Have you tried combining AI with your note-taking system?

🌐 Want to see more of my work? Visit alekdob.com for case studies and detailed project breakdowns from managing software development across 8 European countries.

Clap if this helped you — it helps other developers find these insights! 👏

Keywords: Personal Knowledge Management, Second Brain, Obsidian, Claude Code, AI Productivity, Knowledge Graph, Zettelkasten, Markdown, Enterprise Software, Product Management