Writing / Series

Architecture

Longer-form blueprints. The systems I want to exist, written down with enough rigor to build from.

Some ideas are too big for a short essay. These are the blueprints: retrieval you can actually verify, an Internet of Agents for smart infrastructure, and a marketplace for intelligence that never has to centralize your data. Each one is a system I have argued for, sketched, and in places started to build.

They are written to be built from, not just read. Expect more structure and less story than the other series.

  1. 12 min read VERA: Retrieval You Can Verify A blueprint for a RAG system that earns trust. Hybrid search, a graph of facts, and an agent pipeline that checks its own answer before it ships.
  2. 15 min read The Internet of Agents A blueprint for cities and enterprises where software agents coordinate across silos at machine speed. The architecture, the layers, the rollout.
  3. 11 min read Perspective AI A blueprint for an AI marketplace where data stays owned, models stay diverse, and no single company holds the off switch. Federated intelligence, by design.
  4. 8 min read ZCoP: The Plan That Rewrites Itself A planning doctrine for LLM work. Decompose a goal into a living roadmap, gate every step on a definition of done, and expand the plan lazily as context arrives.
  5. 12 min read Share the Answer, Not the Data In the MCP era your agent negotiates with banks, employers, and hospitals on your behalf. Federated principles decide what it may carry: the answer travels, the data stays home.
  6. 12 min read Sovereign Agents In the agent era, sovereignty stops being about databases and starts being about agency: whose model your agent runs on, whose credentials it carries, and whose instructions it follows. Europe has one of the three already. Here is the whole map.