Why not just use schema.org?
Schema.org is excellent. AI-Me is complementary, not a replacement.
Different scope
Schema.org is a rich vocabulary embedded as JSON-LD across individual pages. It describes entities — products, people, events, organizations — and is consumed by search engines, knowledge graphs, and other crawlers.
AI-Me is a single JSON document at a predictable well-known endpoint. It describes a site as a whole — its canonical pages, offerings, and primary calls-to-action — in a strict, validatable shape designed for AI agents.
Different consumers
Schema.org
- • Search engine crawlers
- • Knowledge graph builders
- • Rich snippet generators
- • SEO tools
AI-Me
- • AI agents
- • LLM-powered assistants
- • Automated discovery tools
- • Agent orchestration platforms
Different trade-offs
Discovery: one fetch vs. crawling
Schema.org data is scattered across pages — an agent must crawl multiple pages to build a complete picture. An AI-Me vendor card is a single document at a known URL. One HTTP request gives agents everything they need.
Shape: strict vs. flexible
Schema.org is deliberately flexible — there are many valid ways to express the same information. AI-Me defines one strict shape with a formal JSON Schema. This makes validation trivial and comparison across sites deterministic.
Purpose: depth vs. discovery
Schema.org excels at rich, detailed descriptions. AI-Me focuses on just enough information for an agent to understand who you are, what you do, and what action to take next.
Use them together
The best approach is to use both. Keep your schema.org markup for search engines, and add a vendor card for AI agents. Link them via the related.schema_org field in your vendor card:
"related": {
"schema_org": "https://example.com/schema.jsonld"
}
This way, AI-Me acts as the discovery layer, and schema.org provides depth. Agents use AI-Me for quick orientation and can follow the related links for richer data when needed.