The Power of Standards
If you are new to XRTM, start with Getting started first. Come here once you are exporting forecasts, integrating tools, or comparing systems.
This section is a plain-language guide. The normative Forecast Object schemas and compatibility policy live in xrtm-org/governance, with implementation notes in xrtm-org/data.
Standards are the bedrock of interoperability. In the fragmented world of AI systems, xrtm establishes a strict contract for how intelligence is shared, evaluated, and improved.
Why We Need a Standard
Without a shared standard:
- Evaluation is Impossible: You cannot compare AI systems if they output different formats.
- Collaboration is Friction: Tools cannot chain together if they don't speak the same language.
- Progress is Stalled: We spend time writing parsers instead of improving intelligence.
The Forecast Object Standard is our answer. It is a rigorous, schema-enforced JSON structure that guarantees every prediction comes with a timestamp, a reasoning trace, and calibrated probabilities. When you need a machine-checked answer, validate against the governance JSON schema instead of treating this page as the canonical contract.
Interoperability
By adhering to this standard, any tool in the xrtm ecosystem can talk to any other. You can swap out the inference engine (xrtm-forecast) for your own custom model, and as long as it outputs a valid Forecast Object, xrtm-eval can score it instantly.