Haem.io

Haem.io Methodology

Haem.io is designed as a deterministic clinical decision-support system for AML and MDS pathways. The core objective is reliable execution of guideline logic with transparent evidence trails, explicit rule provenance, and clinician-verifiable data points.

Method principles

  • Guideline-aligned rule execution for WHO 2022 and ICC 2022 framing.
  • Deterministic outputs for equivalent inputs to reduce interpretation drift.
  • Traceability of outputs to criteria-level evidence and branching logic.
  • No black-box decision layer: input values, criteria branches, and rule outcomes are visible for review.
  • Separation of decision support from autonomous clinical judgment.

Clinical scope

Current production focus is AML and MDS support. Coverage evolves over time, and roadmap pages should be treated as forward-looking product intent rather than immediate clinical capability.

Related source documents

Source docs hubValidation and evidenceNeurosymbolic implementation article

Last updated: March 2026. This content is for healthcare professional decision support and does not replace clinical judgment.