Get started

Introduction

What Metera is, and what it guarantees.

An AI agent that acts on data needs to know whether that data is true. Metera sits between your agent and the sources it would otherwise call directly. For every request it selects a source, runs the response through a chain of checks, and returns the datum together with a verification block stating which checks passed and at what assurance level.

The engine behind this is M1. It scores sources continuously on availability, freshness, schema conformance, and divergence from consensus, and routes each call to the source most likely to deliver. Routing is decided by measurement and by your policy, never by price, payment, or a commercial relationship.

What Metera is not: it does not produce data, it is not an oracle, and it does not guarantee correctness. What it guarantees is that it will not overstate what it knows. When no source can confirm a datum to the level you asked for, it refuses rather than deliver something unverified. That refusal is the product working, not failing.