Profiles

GEO, AI Crawlers, AI Agents, Persistent Memory, M2M (Machine-to-Machine) Architecture

I spent my early career running brand campaigns for companies like Sony, but since 2010 I’ve been in the trenches as a Product Owner building digital businesses.
When generative AI hit the market, I saw two structural problems in how it interacts with WordPress:

Problem 1: AI models scrape messy HTML and hallucinate about brands. Nobody was solving this at the root. So I built LLM Override — a plugin that intercepts GPTBot, ClaudeBot, and 25+ AI crawlers, bypassing the human frontend to serve them clean, exact M2M Markdown. It forces LLMs to ingest the narrative you define, not the garbage code they scrape.

Problem 2: Every AI conversation about your WordPress site starts from zero. Whether you use Claude, ChatGPT, or any agent via MCP — the AI forgets everything between sessions. Your site context, your plugin history, your past decisions: gone. So I built LolaCore — an AI consultant that lives inside your wp-admin, scans your site automatically, remembers everything across sessions, and manages your WordPress through natural language. You bring your own API key (Anthropic, OpenAI, or DeepSeek). Your data never leaves your server.

Both plugins exist because WordPress needs an infrastructure layer for the age of AI agents. That’s what I build.