Neuronautix ← All presentations

Data First

FAIR-by-Design Standardization for NAM Evidence

Damien Huzard, PhD · Neuronautix
11 May 2026 · 12 min

The gap

NAM science is advancing faster than NAM data infrastructure

Assays improved. Interoperability did not.

The bottleneck is metadata, not assay novelty

Current state

Context trapped in prose

Lab-specific endpoint semantics

Legacy data hard to recover

Consequence

Slow, manual reuse

Costly cross-study synthesis

Fragmented AI training corpora

Regulatory framing

Regulatory framing is data-intensive

Context of use, technical characterization, and fit-for-purpose are fundamentally metadata requirements.

FDA draft NAM guidance, 2026

Design-time FAIR beats post-hoc FAIRification

Post-hoc

Expensive reconstruction

Provenance gaps

Uneven record quality

By design

Required fields at source

Validation at ingest

Immediate reuse readiness

Domain model

NAMO provides a practical domain model

Schema-first vocabulary for NAM systems, evidence context, and interoperability.

Deployment

Minimal deployment pattern

pipeline.txt 
Assay execution
  -> Structured capture form
  -> Schema + term validation gate
  -> Versioned evidence record
  -> Reuse query and reviewer export

AI role

AI assists curation; validation enforces trust

LLMs can draft mappings, but deterministic schema checks and human review decide acceptance.

Standardize now.
Accumulate reusable evidence assets — or accumulate remediation debt.

Next step: one 90-day pilot

Select one assay family

Define minimum metadata contract

Enforce validation before acceptance

Pilot historical-control reuse audit

White paper  ·  PDF  ·  Contact

1 / 10