Insights
Long-form thinking on AI, software, and the shift from coding to orchestration. From the inside, not from a distance.
Why Single Agents Hit the Wall
Why Single Agents Hit the Wall. Dissecting the FullStack-Agent Architecture. Benchmark Results: Beyond Proof of Concept.
Why Most AI Pilots Fail — And What To Do Instead
Most AI pilots stall because companies treat them as technology experiments instead of business-outcome problems. Here are the three patterns that separate the ones that ship from the ones that don't.
Why Agents Kill the Per-Seat Model
Why Agents Kill the Per-Seat Model. Case Studies: Decline, Pivot, and Rise. What Actually Works: Building AI-Native Stacks.
What MCP and A2A Actually Solve
What MCP and A2A Actually Solve. The Head-to-Head Reality Check. Production Hybrid Stacks That Actually Work.
The Trust Gap: Where Velocity Meets Reality
The Trust Gap: Where Velocity Meets Reality. Agent Swarms in Production: Beyond the Single-Agent Ceiling. Context: The Real Bottleneck in AI Coding.
The Technical Stack That Delivers Sub-500ms Latency
The Technical Stack That Delivers Sub-500ms Latency. Platform Matrix: What Works for Nordic Enterprise Deployment.
The Tale of Two Protocols: What They Actually Do
The Tale of Two Protocols: What They Actually Do. Production Reality: What Actually Works (And What Breaks).
The Slop Deluge: When Machines Code at Machine Speed
The Slop Deluge: When Machines Code at Machine Speed. When Saying No Becomes Your Superpower. The Velocity Paradox: Faster Code, Slower Systems.
From SaaS Bloat to AI-Native Factories
From SaaS Bloat to AI-Native Factories. Inside the Software Factory: From Vibe Prompts to Production.
S-Tier: Production-Grade Orchestration
S-Tier: Production-Grade Orchestration. A-Tier: Rapid Prototyping Champions. B-Tier: Specialized Excellence.
We're Building an MCP Server for Meeting Transcripts. Here's Why It Changes the LLM Wiki Workflow.
Karpathy's LLM wiki pattern needs raw sources you control. Every meeting tool launched a read-only MCP server in the last six months. We're building the one that closes the loop.
The Protocol Stack That Changed Everything
The Protocol Stack That Changed Everything. From Single Agents to AI Engineering Teams. The A.G.E.N.T. Playbook: A Builder's Guide to 10x Gains.
The Protocol Landscape: MCP vs A2A Fundamentals
The Protocol Landscape: MCP vs A2A Fundamentals. The 25,000-Task Revelation: Sequential Beats Hierarchical.
The Productivity Paradox: When Faster Generation Hits Reality
The Productivity Paradox: When Faster Generation Hits Reality. The Science of Human-AI Collaboration: Understanding the Novelty Bottleneck.
The Numbers Don't Lie: When Velocity Becomes Paralysis
The Numbers Don't Lie: When Velocity Becomes Paralysis. Where Traditional Workflows Break Down. Emerging Solutions: Beyond Human-Scale Review.
The Multi-Agent Coordination Crisis
The Multi-Agent Coordination Crisis. CAID: Software Engineering Primitives for AI Teams. The Numbers: Why Coordination Beats Compute.
The Multi-Agent Moment: Why 2025 Changed Everything
The Multi-Agent Moment: Why 2025 Changed Everything. MCP: The Universal Tool Interface That Actually Works.
19 MCP Tools for AI Memory: What MemPalace's Read-Write Pattern Teaches Meeting Builders
MemPalace shipped 19 MCP tools that include knowledge graph mutation, agent diary writes, and auto-dedup. Most memory MCPs are read-only. Here's what the read-write pattern looks like — and why meeting tool builders should be paying attention.
The Mechanics of Destruction: How Agents Kill SaaS
The Mechanics of Destruction: How Agents Kill SaaS. The Carnage: Winners, Losers, and the New Hierarchy. The New Software Stack: From Seats to Outcomes.
MCP: The USB-C Port for AI Agents
MCP: The USB-C Port for AI Agents. A2A: When Agents Need to Talk to Each Other. The Architecture Decision: MCP + A2A or Pick One?.
MCP: The Universal Adapter for AI Tool Access
MCP: The Universal Adapter for AI Tool Access. A2A: Orchestrating Agent Swarms Like Human Teams. The Architecture of AI Engineering Teams.
MCP: The Universal Tool Adapter That Actually Works
MCP: The Universal Tool Adapter That Actually Works. A2A: Multi-Agent Orchestration for Enterprise Scale. The Hybrid Stack: Why Smart Builders Use Both.
MCP: The Universal Tool Adapter
MCP: The Universal Tool Adapter. A2A: The Language of AI Teamwork. Head-to-Head: When to Choose What.
MCP: The Tool Whisperer
MCP: The Tool Whisperer. A2A: Agent Diplomacy at Scale. Production Reality: What Actually Works.
MCP: The Standard That Makes AI Agents Actually Useful
MCP: The Standard That Makes AI Agents Actually Useful. A2A: When AI Agents Need to Talk to Each Other.
MCP: The Precision Instrument
MCP: The Precision Instrument. A2A: The Collaboration Engine. The Hybrid Reality: Best of Both Worlds.
MCP: The Internal Wiring Standard
MCP: The Internal Wiring Standard. A2A: The Coordination Layer. The Hybrid Architecture Reality.
MCP: Giving Agents Access to Everything They Need
MCP: Giving Agents Access to Everything They Need. A2A: Agents That Actually Collaborate. How MCP and A2A Work Together: The Full Stack.
The MCP Foundation: Agents That Actually Use Tools
The MCP Foundation: Agents That Actually Use Tools. A2A: The Delegation Layer for Agent Teams. The Hybrid Architecture: MCP + A2A in Production.
MCP: The Foundation Layer That Actually Works
MCP: The Foundation Layer That Actually Works. A2A: When Agents Become Colleagues. The Stack That Scales: Why MCP + A2A = Magic.
The MCP Foundation: Vertical Integration That Actually Works
The MCP Foundation: Vertical Integration That Actually Works. A2A: Horizontal Orchestration for Enterprise Reality. The Power of Protocol Convergence.
How to Build a Karpathy-Style LLM Wiki From Your Meetings (Obsidian Tutorial)
A practical Obsidian workflow for piping meeting transcripts into Karpathy's LLM wiki pattern. Two paths: manual export and an MCP server. Includes the schema, prompts, and example queries.
The Hidden Cost of Free Code
The Hidden Cost of Free Code. Architecture Emerges as the New Bottleneck. The Great Developer Evolution.
The Great Productivity Paradox: More Code, Same Velocity
The Great Productivity Paradox: More Code, Same Velocity. From Coding Bottlenecks to Context Gaps. The Rise of Developer-Orchestrators.
The Great Commoditization: When Implementation Becomes Free
The Great Commoditization: When Implementation Becomes Free. The Three Judgment Bottlenecks Killing AI-Native Teams.
The Great Code Shift: Volume Up, Quality Down
The Great Code Shift: Volume Up, Quality Down. Why AI Code Fails: The Judgment Gap. What Works: Lessons from the Frontier.
The Framework Wars: Choosing Your Organizational Structure
The Framework Wars: Choosing Your Organizational Structure. Benchmarks as OKRs: Measuring Team Performance. Protocols: The APIs of Agent Communication.
The Extinction Evidence: Why Traditional SaaS Is Hemorrhaging Value
The Extinction Evidence: Why Traditional SaaS Is Hemorrhaging Value. The New Breed: Y Combinator's Agent-Native Pivot.
The Evidence: Real Companies Making Real Swaps
The Evidence: Real Companies Making Real Swaps. From Vibe-Coding to Production: Anatomy of AI-Native Software.
The Evidence: AI Code Quality Is Worse Than We Admitted
The Evidence: AI Code Quality Is Worse Than We Admitted. The New Bottleneck: Systems Thinking vs. Syntax Generation.
The Economics of Custom Everything
The Economics of Custom Everything. Production-Ready Agent Frameworks. Benchmark-Proven Reliability.
The Economics of Extinction
The Economics of Extinction. Production-Ready Playbook: From Audit to Scale. Case Studies: Where Agents Are Winning.
The Demo That Changes Everything
The Demo That Changes Everything. The Gartner Prediction: SaaS Becomes Infrastructure. The Reality Check: When AI-Built Apps Meet Production.
The Death of the Dashboard
The Death of the Dashboard. Real-World Replacements Happening Now. The Nordic Advantage: Building Agent-Native.
The Commoditization Evidence: When Syntax Becomes Worthless
The Commoditization Evidence: When Syntax Becomes Worthless. Agentic Codebases: The Infrastructure Revolution.
From Chatbots to Full-Stack Builders
From Chatbots to Full-Stack Builders. What Production-Ready Agent Apps Actually Look Like. The Nordic Advantage: Open Source and Pragmatic Implementation.
From Chaos to Coordination: Why Agent Sprawl is Killing ROI
From Chaos to Coordination: Why Agent Sprawl is Killing ROI. The Protocol Wars: MCP vs A2A and Why You Need Both.
From Bash Scripts to AI CTO
From Bash Scripts to AI CTO. Architecture: The Orchestration Layer. The Self-Build: 8 Days of Autonomous Development.
The Architecture That Actually Makes Sense
The Architecture That Actually Makes Sense. Why This Matters for Your Engineering Team Right Now. Production Frameworks: What's Actually Working in 2026.
The Architecture of AI Coordination
The Architecture of AI Coordination. When Micromanagement Wins: The MCP Sweet Spot. The Delegation Advantage: A2A in Practice.
The Anatomy of an AI Software Agency
The Anatomy of an AI Software Agency. From Installation to Production: A Builder's Walkthrough. Real-World Demos: When AI Teams Ship Software.
Is Teamtailor Compliant with the EU AI Act?
Teamtailor is one of the most popular ATS platforms in the Nordics. If you use its AI features to screen or rank candidates, it's classified as high-risk under the EU AI Act. Here's what that means for you.
Is Intercom Compliant with the EU AI Act?
Intercom's Fin AI chatbot is limited-risk under the EU AI Act — but you still need to disclose that users are talking to AI. Here's what the law requires and how to get it right.
Is HireVue Compliant with the EU AI Act?
HireVue's AI-powered video interviews are the textbook example of high-risk AI under the EU AI Act. If you use them, you have serious compliance obligations. Here's exactly what they are.
EU AI Act Compliance for HR Teams: A Practical Guide
HR is the single highest-risk area under the EU AI Act, and most People teams don't know it yet. This guide covers what HR directors need to do — from candidate screening to performance reviews — before enforcement begins.
What the EU AI Act Means for SaaS Companies
SaaS companies face a unique challenge under the EU AI Act: you might be both a deployer of AI tools and a provider of AI systems. This guide explains the difference, why it matters, and what CTOs need to do now.
AI Act for Swedish Companies: What You Need to Know
Sweden has some of Europe's highest AI adoption rates but dangerously low awareness of the EU AI Act. This guide covers what Swedish business leaders need to know — including the role of IMY, common Swedish tools, and why there is no grace period.
Is Gong Compliant with the EU AI Act?
Gong is fine for sales coaching and deal intelligence. But if you use it for employee performance evaluation, it becomes high-risk under the EU AI Act. The line is thinner than you think.
Is GitHub Copilot Compliant with the EU AI Act?
Good news: GitHub Copilot is minimal to limited risk under the EU AI Act. But there's one obligation every company using it must meet — AI literacy under Article 4.
The EU Digital Omnibus: What the Proposed Delay Means for AI Act Compliance
The European Commission's Digital Omnibus package proposes pushing high-risk AI deadlines to December 2027. But transparency and literacy obligations remain unchanged — and they're already here.
AI Act Article 4: The Obligation That's Already Active (And What It Means)
Article 4 of the EU AI Act — the AI literacy obligation — has been in force since February 2025. Here's what it actually requires, what 'sufficient literacy' means in practice, and how to implement it.
5 AI Systems Your Company Probably Uses That Are High-Risk Under the AI Act
Most companies don't realise they're already deploying high-risk AI systems. Here are five common tools — from recruitment screening to customer eligibility chatbots — that trigger full AI Act compliance obligations.
The Agyn Architecture: Production Multi-Agent Done Right
The Agyn Architecture: Production Multi-Agent Done Right. CAID: Git-Native Async Delegation That Scales. Orchestration Patterns That Actually Work.
The Agent Coordination Problem We're Solving
The Agent Coordination Problem We're Solving. MCP: The Deep Dive Protocol for Agent-Tool Integration. A2A: Enabling True Multi-Agent Collaboration.
The Agent Arsenal: Production-Ready Powerhouses
The Agent Arsenal: Production-Ready Powerhouses. The Three Paths to SaaS Obsolescence. What "Good" AI Software Actually Looks Like.
The 25,000-Task Reality Check
The 25,000-Task Reality Check. The Endogeneity Paradox: Why Structure Kills Performance. Where Popular Frameworks Go Wrong.
The Build-vs-Buy Equation Has Flipped
The Build-vs-Buy Equation Has Flipped. Real-World Replacements: The Evidence Mounts. Market Dynamics: Winners, Losers, and the Great Unbundling.
The Triggers: AI Agents and the Vibe Coding Revolution
The Triggers: AI Agents and the Vibe Coding Revolution. The Existential Crisis: Why SaaS Valuations Are Crumbling.
Inside the Study: Architectures, Benchmarks, and Shocking Results
Inside the Study: Architectures, Benchmarks, and Shocking Results. The Three Scaling Laws: Predictable Patterns in Agent Swarms.
Want to go deeper?
We explore the frontier of AI-built software by actually building it. See what we're working on.