Notes
Writing on AI operations, enterprise architecture, and what it actually takes to deploy artificial intelligence inside a real organization.
-
May 7, 2026
AI Deployment Case Studies
An anonymized set of AI deployment case studies from real operating work across sales, marketing, inventory, fulfillment, and executive routines. Start with the work, not the model.
-
April 17, 2026
Living ICPs: Making Customer Profiles Actually Useful
Most companies treat ICPs as static decks. Ours are live — auto-refreshing, conviction-scored, and always feeding the next test. Here's the actual pipeline.
-
April 12, 2026
Building AI Operations in Reality
What it actually looks like to deploy AI agents in a real company. No theory. Just the parts (runtime, fleet, shared brain, channels), how they work together, and what I learned trying to make bots productive employees. Discord for the work floor. Telegram for the exec layer.
-
April 12, 2026
The Systems Flywheel
How AI operations get better with every deployment — and why each implementation makes the next one faster, cheaper, and more valuable. The real power isn't in the tool. It's in the methodology flywheel.
-
April 11, 2026
The Ad Machine
We built an ad generation system in a day — something that reads performance data, compiles creative briefs, generates Seedance 2.0 videos, and posts them for review every Monday. Here’s what we learned building it.
-
April 11, 2026
The System That Gets Better by Being Used
Most automation systems are static. Here’s what actually makes a system self-improving — and why the answer isn’t machine learning.
-
April 11, 2026
Where Do the Bots Live?
One of the first real questions you hit when deploying AI agents inside a company is deceptively simple: where do they live? Not technically — operationally.
-
April 7, 2026
The Control Problem: Managing 4 Brands with AI Agents
When you run multiple product lines with a small team, the bottleneck isn't talent or budget. It's coordination. Here's how I'm building a system where the agents do the status tracking and humans just talk.
-
March 28, 2026
The Business Brain: How Specialized AI Agents Mimic Neural Systems
A fleet of specialized AI agents, each monitoring a business domain, all feeding a shared database. That's not a tech stack. That's a nervous system.
-
March 28, 2026
The Operating Rhythm Problem: How We Built a Weekly System That Runs Itself
Every operating framework fails at small companies for the same reason: the maintenance burden falls on one person who's already underwater. Here's how we built our own operating rhythm with AI handling the upkeep — and why email reply beats a dashboard login.
-
March 15, 2026
Building Instead of Buying: Two Internal Tools We Built With AI Coding Agents
We didn't set out to disrupt SaaS. We had operational problems and coding tools. Here are two internal projects we built that probably would have been SaaS subscriptions two years ago.
-
March 1, 2026
The Operations Bot Design Framework
Most automation fails not because the technology is wrong but because the design is wrong. Eight principles for building bots that actually run reliably in production.
-
February 28, 2026
The Lattice: How Multi-Agent Systems Compound Intelligence
Most AI deployments are single-agent. What I'm building is different — a network of specialized agents that share memory, produce signals for each other, and compound in value the longer they run.
-
February 27, 2026
The Outbound Machine: Building a B2B Pipeline That Gets Smarter Every Week
Custom landing sites, a swarm of specialized agents, marketing psychology at scale, and a feedback loop that updates its own angle weights from real reply data. Here is the full architecture.
-
February 27, 2026
The Digital Agency in a Box
Most companies outsource their operational intelligence to agencies — one for paid media, one for content, one for vendor management, one for social. Four retainers, four communication styles, and you playing coordinator. What if that entire layer could be autonomous?
-
February 21, 2026
We Stopped Building Product Pages. We Built a Factory Instead.
Generic product pages don't convert niche buyers. So we built a monorepo that produces persona-specific landing pages from a single template — with full-funnel tracking, multiple checkout paths, and 3-day turnaround. Here's the architecture, and we're open-sourcing the template.
-
February 19, 2026
Every Company Needs a Chief AI Officer. Most Don't Know What That Means.
AI projects fail at the organizational layer, not the technology layer. No one owns it. CTO treats it as IT. CMO uses it for copy. CEO watches demos. Meanwhile competitors are operationalizing it. Here is what building an AI system actually requires — and what the person who does it well looks like.
-
February 10, 2026
Building the Operations Brain: How I Architected an AI Fleet That Runs My Business While I Sleep
AI agents are everywhere. Production AI operations systems are nowhere. How I architected a fleet that runs business operations across time zones — while I sleep.
-
February 8, 2026
Designing Systems That Learn: Fleet Bots and Self-Recycling Memory
How I deployed three specialized bot fleets with memory systems that get smarter over time. It's not about what AI can do today — it's about designing systems that generate their own intelligence.
-
February 7, 2026
Building a Fleet of AI Agents for Enterprise Operations
How I deployed 5 specialized AI agents across a public company's operations using OpenClaw. Division-based architecture, human-in-the-loop design, and why specialists beat generalists.
-
January 26, 2026
Day One with Clawdbot
I spent a full day tinkering with a system that runs on my server. Not an app. Not a dashboard. A terminal. AI doesn't replace presence. It creates the conditions for it.
-
January 26, 2026
The Terminal is the New Interface
GUIs are going extinct. Raw thoughts in, action out. A partner that understands you.
-
January 2026
Building an AI Operating System for a Public Company
How I built 10 specialized AI agents to run operations across 6 departments, automating everything from Monday morning briefings to board report compilation. Architecture, code patterns, and real results.