Reasoning as Code & governed AI

Essays on the missing layer between enterprise data and AI agents—playbooks, proof, ReBAC, and shipping production copilots that audit themselves.

Can ontology AI predict the FIFA World Cup?

Every tournament, someone asks an AI for a winner and gets confident fluff. Ontology AI is not a crystal ball — it is governed structure across squads, fixtures, and injuries. What it can and cannot predict, and why that matters for agents and enterprises alike.

Rethinking data in the AI age

From SQL joins to NoSQL, graph DBs, and lakes — each solved real problems. Agents need something different: a real-time reasoning layer like AnythingGraph that stores playbooks, not rows, and queries where data already lives.

AnythingGraph is a new database — and it stores no data

In the AI age, your rows already live in Postgres, Salesforce, and CSV files. What agents need is governed meaning — playbooks, bindings, access, and proof — not another copy. Why a reasoning database matters now.

The surprising link between UAP, ontology, and AI

Unidentified phenomena, machine learning, and formal ontologies look unrelated — until you see they all solve classification under uncertainty. What NASA’s UAP report and enterprise playbooks have in common.

Let Claude build your employee payroll playbook via MCP

Connect Claude to Anything CLI, introspect Postgres and CSV sources, and auto-create the crm-payroll-access playbook — with real MCP prompts, tool calls, and saved JSON/YAML.

How to check available data sources via Anything Graph MCP in 2 minutes

Siloed CRM, payroll, and SaaS data leaves agents blind. AnythingGraph lists every connected source via MCP, introspects schema in place, and federates silos in playbooks — no ETL required.

Why we no longer want to move data around in the AI age

ETL, lakes, and vector copies made sense for dashboards — not for agents asking every minute. Why query-in-place, governed playbooks, and proof beat another data migration project.

From playbook to compliance verdict in minutes

Step-by-step: install the Fintech transaction compliance playbook, explore the graph, connect Claude via MCP, and ask whether transaction TX-8842 can be approved — with an explainable answer.

Top modern AI use cases for ontology

Why ontology is back in the age of AI — and why cybersecurity, finance, insurance, healthcare, manufacturing, and supply chain need shared meaning, not just better summaries.

The missing layer between your data and your AI agents

Why RAG and MCP are not enough, what Reasoning as Code means, and how AnythingGraph lets developers version ontology, access, and proof in playbooks—then expose them to any agent via governed MCP.