I Do Run AI Agents Overnight. Here's What Actually Matters.
Boundaries, supervision, evidence, and why unattended execution is not the same thing as unbounded autonomy.
Petie Clark — Applied AI Systems
Applied AI systems that survive contact with the business.
Field notes on agent orchestration, evaluation, authority boundaries, enterprise integration, deployment, and production operations.
Get new posts in your inbox
No spam. No marketing. Just the writing.
/start-here
How I think about governed systems
Why I Prefer One Clear Source of Truth
Why tools can multiply while authority, state, and accountability still need one dependable home.
Why I Write Down the Rollback Before the Change
A deployment is not ready merely because the forward path looks convincing.
The Case for a Smaller Blast Radius
Production systems become trustworthy through bounded change, not theatrical confidence.
The Thin Control Plane Is Usually Enough
A useful control plane coordinates intent and evidence without becoming a second execution engine or a counterfeit source of truth.
4 min read
Read →
Local-First Is an Operating Choice, Not a Lifestyle Brand
Keeping a large runtime local only works when the machine, storage, and recovery behavior are operated as one system.
4 min read
Read →
What Restore Testing Teaches You About Every Other System
The backup report said the archives were valid. The restore diary began because valid archives weren't the claim I needed to prove.
4 min read
Read →
The Useful Part of Observability Starts After the Alert
An alert earns its keep when it shortens the path from a symptom to a safe decision.
4 min read
Read →
I Leave Capacity for the Weird Failure
Decision: routine workloads will not consume the space needed to preserve, copy, and compare a failed state. Capacity planning will include a diagnostic maneuver, not just normal demand and restart overhead.
4 min read
Read →
Every Service Adds a Future Chore
A service doesn’t add one generic maintenance task. It adds work at every dependency boundary where its assumptions can diverge from the next version underneath it.
4 min read
Read →
Why I Keep the Old Path Until the New One Proves Itself
Replacing a scheduled path is deceptively simple. The old job runs on a clock, the new worker reacts to a richer trigger, and both appear to perform the same task. It’s tempting to disable the old entry, enable the new one, and call the migration complete when the first run succeeds.
4 min read
Read →
When Convenience Creates a Second System
A container management screen can make one small change feel harmless. Select a running service, adjust an environment value, click redeploy, and the new behavior appears. The repository remains untouched. No syntax to remember, no file to open, no review needed.
4 min read
Read →
The Useful Difference Between Reachable and Ready
A listening port is a wonderfully low bar. It proves that something accepted responsibility for an address. In local model infrastructure, that can happen well before the system is capable of doing the job that brought a caller there.
4 min read
Read →
The AI Productivity Paradox Is Real - But You Are Measuring the Wrong Thing
CEOs say AI isn't moving the needle on productivity. I've been running AI in production for months. Here's why they're both right and wrong — and what the actual measurement problem is.
5 min read
Read →