Insights
Field notes on AI systems,
GTM, and revenue architecture.
Practical breakdowns from systems I build and run. Not theory: architecture, stack choices, and operator lessons from live deployments.
The Claude Code GTM Agent Starter Pack
The Claude Code GTM Agent Starter Pack
How I built a four-layer AI system that runs outbound research, drafts sequences, and enriches CRM records: without a single external enrichment tool.
- A four-layer Claude Code agent stack handles outbound research, sequence drafting, CRM enrichment, and ICP scoring. No Clay. No enrichment credits. No n8n flows.
- 50-account enrichment run: 75 minutes total (45 unattended + 30 review) vs. 3-4 hours with a Clay + copywriting workflow.
- The system runs at 5-10% of equivalent Clay cost, with better output quality on the reasoning and copy layers.
20 articles
Latest
The browser bookmark is the wrong primitive
Why I built Kade as a local-first link vault: not another bookmark list, but a private queue for links that need context before they are opened again.
How I use Claude Code in client GTM work
Beyond the outbound agent stack: how Claude Code runs content operations, CRM hygiene, and research inside real client engagements: what the delivery model looks like from the outside.
The AI stack I actually run in production
A complete tour of every AI tool in active use across my client work and products: what earned its place, what got cut, and what the operating logic looks like across the stack.
A GTM agent that runs outbound, enriches CRM, and costs $2 per run
A GTM agent that runs outbound, enriches CRM, and costs $2 per run
No Clay, no Apollo credits, no n8n. A four-layer Claude Code agent that handles outbound research, sequence drafts, and CRM enrichment natively.
The browser bookmark is the wrong primitive
The browser bookmark is the wrong primitive
Why I built Kade as a local-first link vault: not another bookmark list, but a private queue for links that need context before they are opened again.
How I use Claude Code in client GTM work
How I use Claude Code in client GTM work
Beyond the outbound agent stack: how Claude Code runs content operations, CRM hygiene, and research inside real client engagements: what the delivery model looks like from the outside.
The AI stack I actually run in production
The AI stack I actually run in production
A complete tour of every AI tool in active use across my client work and products: what earned its place, what got cut, and what the operating logic looks like across the stack.
Why CRM-first beats prompt-first in AI adoption
Why CRM-first beats prompt-first in AI adoption
Most B2B AI adoption starts with prompts and ends in chaos. The entry point isn't the tool: it's the data layer. Here's why CRM hygiene is the most important AI project most companies aren't running.
MDX components showcase
MDX components showcase
Living reference for rich MDX components available in insights articles.
From Framer to Astro: the build vs buy decision behind this site
From Framer to Astro: the build vs buy decision behind this site
The honest trade-offs behind rebuilding wojciech.io: why I moved from a visual builder to Astro, what MDX and i18n changed, and which debt I accepted on purpose.
Cloudflare migration: what you get beyond hosting
Cloudflare migration: what you get beyond hosting
Moving to Cloudflare Pages is straightforward. What most people miss is the security and identity stack that comes free with it: Zero Trust, Access, Tunnel, and WAF.
Rebuilding my site with Claude Code, Codex, and zero frontend experience
Rebuilding my site with Claude Code, Codex, and zero frontend experience
The full story of migrating from Framer to Astro + Cloudflare Pages using AI agents. What I directed, what they built, and where the human loop actually mattered.
macOS Teleprompter for MacBook Notch: building a native Swift app
macOS Teleprompter for MacBook Notch: building a native Swift app
How I built a native macOS teleprompter that uses the MacBook notch to keep speaker notes close to the camera without breaking eye contact on calls.
What 10+ shipped micro-apps taught me about building with AI
What 10+ shipped micro-apps taught me about building with AI
AI makes shipping fast. Distribution stays hard. After 10+ micro-SaaS apps, here is what I learned about product decisions and the limits that actually matter.
I replaced the Google Ads UI with a dashboard that tells me what to do next
I replaced the Google Ads UI with a dashboard that tells me what to do next
The native Google Ads interface is optimized for Google's revenue, not for operator decisions. So I built AdsAI: budget, ROAS, ICP fit, and next actions in one view.
Four layers that make a GTM agent actually work in production
Four layers that make a GTM agent actually work in production
Operator identity, MCP tools, client memory, reusable skills. Here is what the architecture looks like when the agent runs real GTM work, not a demo.
When to build your GTM tool and when to just buy one
When to build your GTM tool and when to just buy one
The danger zone is buying SaaS for something that should be a workflow and building for something a $50/month tool already handles. Here is how I decide.
Most AI adoption frameworks are built for demos, not for teams
Most AI adoption frameworks are built for demos, not for teams
Growth teams try AI, nothing sticks, six months later they are back to the old workflow. The problem is not capability. It is integration. Here is the framework I use instead.
Campaigns plateau. Systems compound. The difference matters.
Campaigns plateau. Systems compound. The difference matters.
Growth teams hit a ceiling because they run campaigns when they should be building systems. Here is what a revenue system looks like when it actually works.
Your CRM is an expensive address book. Here is how to fix that.
Your CRM is an expensive address book. Here is how to fix that.
Most B2B CRMs collect contacts and produce reports nobody reads. Here is how I rebuild them into revenue systems that run without daily manual input.
Building a booking engine for a market nobody else wanted
Building a booking engine for a market nobody else wanted
Niche campsite operators needed calendar, pricing, and SMS. Generic platforms over-engineered everything else. So I built what they actually needed.
The growth system I built when founder-led sales stopped scaling
The growth system I built when founder-led sales stopped scaling
Founder closes deals until they can not. Then you need a system. This is the growth infrastructure I built from ICP clarity through acquisition to retention.
No articles in this category yet.
Try a different filter.
Newsletter
Want to know more first?
New articles on AI systems, GTM architecture, and growth operating models, delivered when they publish.