Tag: systems
All the articles with the tag "systems".
How I Built an AI Harness That Ships Products on Its Own
Updated: at 08:34 PMEveryone is waiting for a smarter model. I spent that time building the operational wrapper around the model we already have, a multi-agent harness that takes a one-line task and returns a deployed product for about five cents. Here is how it actually works, and what shipping hundreds of autonomous tasks taught me.
The Inference Substrate, or Why Every AI App Pays a Tax It Cannot See
Updated: at 11:26 AMEvery company building on closed AI models pays a permanent, growing tax in cost, latency, and wasted generality. There is no infrastructure layer fixing it, and the model companies structurally cannot build one. This is the thesis for the layer that should exist in the empty middle, and why it is an infrastructure and cryptography problem rather than an AI research one.
A Coding Agent That Proves Its Fixes and Never Leaves Your Network
Updated: at 02:08 PMA bank cannot send its source code to a US AI API. So I built a coding agent that runs fully air-gapped on the customer's own hardware, is model-agnostic, and never claims a fix works until it has run the repo's own tests and watched them pass. The interesting result was that the harness, not the model, was where the capability lived.