Demo agents fall over in production because errors compound across steps and prompts break when the business process changes. Structured output, LLM-as-a-judge evals and OKR alignment are what get a prototype across.
AI now generates faster than any of us can type, which makes judging the output the actual job. Where LLM code tends to break in WordPress: hooks that no longer exist, uncached get_option calls in loops, and raw SQL that HPOS will punish.
Most machine learning projects fail on framing, not on model quality. Five checks to run before you open a notebook: name the decision, price the two kinds of error, audit the target variable, simulate the deployment, and write down how the project fails while the metrics look fine.
WordPress 7.0’s AI updates are heading into Beta 2. This post covers the pivot away from the Abilities API, the new Connector Screen, and the dependency issues in the AI Experiments plugin.
February was short but it made a few things obvious: solo work stalls ML projects, code is not documentation, and an MLOps stack that ignores the deployment environment never ships. Notes from a month of ML and backend work.
Most AI features never leave the notebook. Notes on the production data architecture that gets them into a live site: separated layers, database integrity over raw insert speed, and the state and permission problems agents bring with them.
Gemini’s open-vocabulary detection removes the label-thousands-of-images step from a vision pipeline. Notes on structured JSON output with Pydantic, editing detected regions with Nano Banana, and the rate limits that show up in production.
WordPress.com built its new AI Assistant directly into the block editor and Media Library. It edits blocks and generates images with Google’s Gemini models, though full layout control needs a Block Theme.
Research on 13 LLMs in simulated markets found the agents holding price floors nobody asked them to hold. For a WooCommerce store running an AI pricing plugin that is an antitrust problem, and the fix sits in how the pricing logic is written.
Claude Code pays off when you build it tools instead of asking it for snippets. A CLAUDE.MD map, real WP-CLI commands, and a small plugin audit script turned a four-hour debugging session into a check that runs in seconds.