The AI Architect

The AI Architect

Home
Podcast
Notes
Archive
Leaderboard
About
The AI Chatbot Era Is Ending. Teams Are Optimizing the Wrong Layer.
The next AI skill is not prompting. It is knowing what work to delegate, what authority to give, and what evidence proves the agent finished.
 • Tyler Folkman
Your AI Agent Doesn’t Need Better Prompts. It Needs a Goal.
The new AI skill is not writing better instructions. It is building a harness that remembers the objective, checks the work, and keeps going until the…
 • Tyler Folkman
I Stopped Chasing Better Prompts. These 5 Loops Cut My AI Babysitting.
The reusable systems that made drafts, code reviews, and research passes repeatable without steering every turn by hand.
 • Tyler Folkman
Before You Trust an AI Agent, Make It Fail Locally
The Boring Agent Test: a Microsoft Agent Framework walkthrough for developers who want traces, artifacts, and a workflow they can actually debug.
 • Tyler Folkman
Claude Code Workflows Are Here. Don’t Use Them Like an Intern Swarm.
More agents is not the point. The value is cleaner context, bounded roles, and deterministic review gates.
 • Tyler Folkman
I Was Wasting 31% of My Context Window Before I Fixed My MCP Setup
Loading every tool all the time is the tax.
 • Tyler Folkman
I Asked My AI Agent to Steal My Secrets. It Worked.
I tested the obvious fixes. The one everyone reaches for first is not a security boundary.
 • Tyler Folkman
My AI Agent Burned 1.2B Tokens. A $0 Audit Showed Me Where.
Harness engineering isn't a research paper. It's a weekly habit that reads the logs you already have.
 • Tyler Folkman
I Routed 2,415 AI Agent Turns Across 6 Models. It Cost $76.77
The full router has 8 model paths. Real usage was 94% cheaper than sending every turn to GPT 5.5.
 • Tyler Folkman
I Cut My AI Agent Costs 7x Without Switching Models
rtk, context-mode, and Kimi K2.6 changed the equation. The Uber story shows how expensive agents get when nobody optimizes context.
 • Tyler Folkman
I Built a $0/API Local AI Lab With Two GPUs
Local LLMs are not just about saving money. They are about privacy, experiments, and breaking agents without API anxiety.
 • Tyler Folkman
I Tested Opus 4.7 Against Sonnet 4.6. The Newer Model Lost.
The official benchmarks said Opus 4.7 was better. My own testing says otherwise.
 • Tyler Folkman
I Replaced Claude Code With a $45/Month Multi-Agent System
Claude Code quality dropped 67%. Here's what I built instead.
 • Tyler Folkman
The Claude Code Leak Showed Me What I Was Configuring Wrong
What Anthropic's own code says about how you should actually configure your setup.
 • Tyler Folkman
The AI Architect
The AI Architect
Field-tested AI agent workflows for engineering leaders and builders. I test Claude Code, Codex, local models, and multi-agent systems so you can ship faster without wasting money.
Recommendations
View all 23
The VC Corner
The VC Corner
Ruben Dominguez
Fish Food for Thought
Fish Food for Thought
Mike Fisher
User's avatar
Sara S
Sara S
NEW ECONOMIES
NEW ECONOMIES
Ollie Forsyth
Product Management IRL
Product Management IRL
Amy Mitchell

The AI Architect

AboutArchiveRecommendationsSitemap
© 2026 Tyler Folkman · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture