EAIMS — Enterprise AI Maturity Standard

An open, vendor-neutral, evidence-based, and executable specification for enterprise AI maturity.

Version: 0.2.1 Research Baseline
Status: Public Review and Pilot Preparation
Documentation and assessment content: CC BY 4.0
Code, workflows, and schemas: Apache-2.0

What EAIMS adds

EAIMS combines an enterprise operating model with reproducible assessment infrastructure:

• 9 dimensions, 27 capabilities, and 162 observable maturity anchors
• Capability-specific evidence examples, counter-evidence, and freshness guidance
• Confidence reporting and critical gates that prevent unsafe averaging
• Deterministic Python scoring engine, CLI, schemas, automated tests, and reports
• Three complete fictional assessment fixtures
• Independent conceptual crosswalk to major public AI-governance themes
• Conformance classes, assessor handbook, pilot protocol, and benchmark protocol
• Dependency-free browser reference interface and local Docker deployment

EAIMS supports predictive ML, generative AI, RAG, agents, third-party services, and cloud, hybrid, sovereign, or on-premises architectures. It does not prescribe a vendor, model, platform, or consulting provider.

Project status and stewardship

EAIMS was founded and initially authored by Elias Naserkhaki, registrant and founding steward of eaims.org. Founder attribution records project origin and stewardship; it does not create accredited standards authority or override published governance.

EAIMS is not an ISO, IEC, ANSI, governmental, or accredited standard. It does not provide certification, legal compliance, or safety assurance. A narrowly scoped implementation listing, when present in the public registry, is not certification or approval of customer results. See Disclaimer, IP Policy, Origin, and Trademark Policy.

Quick start

Run the reference engine without third-party runtime dependencies:

python -m pip install -e .
eaims validate examples/fictional-manufacturer.assessment.json
eaims score examples/fictional-manufacturer.assessment.json
eaims report examples/fictional-manufacturer.assessment.json --format html --output report.html

Run the browser interface:

python -m http.server 8080
# open http://localhost:8080/site/

Or use Docker:

docker build -t eaims .
docker run --rm -p 8080:80 eaims

Assessment flow

1. Define scope and evidence handling with the Assessor Handbook.
2. Complete the questionnaire using the evidence catalog.
3. Validate and score the assessment with the CLI.
4. Apply the Conformance claim appropriate to the evidence and review process.
5. Produce findings and a roadmap; never present the result as certification.

Key documentation

Need	Resource
Normative foundation	EAIMS Standard and Capability Matrix
Scoring	Scoring Methodology
Evidence	Evidence Catalog
Facilitation	Assessor Handbook
Implementation claims	Conformance
Approved implementations	Implementation Registry
Research pilots	Pilot Protocol
Future benchmarking	Benchmark Protocol
External frameworks	Independent Conceptual Crosswalk
Change process	RFCs and Decisions

Validation status

v0.2 is executable and internally tested. It has not yet completed independent multi-organization validation, inter-rater reliability research, academic peer review, or representative benchmarking. The repository publishes protocols for generating that evidence honestly.

Contributing and support

Contributions require DCO sign-off and rights disclosure. See Contributing, Reviewer Program, Sponsorship Policy, and Financial Transparency. Funding cannot buy changes, favorable scores, certification, endorsement, or governance control.

Publications

Books and research outputs: EAIMS Publications

Project website

https://eaims.org