Assessment Methodology | Muuvment Labs

Overview

6-Layer Architecture

The assessment is built on six layers that work together to produce a report calibrated to your specific organisation, industry, and jurisdiction. Each layer is independent — if one encounters an issue, the others continue and a report is still produced.

1. Company intelligence

Your website is analysed to extract your company profile, industry sector, geography, and technology signals. This context shapes the entire report.

2. Question engine

17 scored questions across 5 chapters. Each answer maps to a deterministic score anchor — the AI cannot override these scores.

3. Consistency checking

A rules engine checks your answers for logical contradictions before scoring begins, and flags any inconsistencies found.

4. Hybrid retrieval

Relevant context is retrieved from a curated knowledge base using semantic search over 963 embedded document chunks.

5. Regulatory mapping

A live regulatory intelligence engine maps your jurisdiction and industry to applicable regulations — updated in real time at report generation.

6. Multi-model synthesis

Three specialised AI models analyse your data and produce the executive summary, industry context, and strategic recommendations.

Questions

Assessment Chapters

The 17 questions are organised into five chapters. Each chapter probes a distinct organisational capability that determines AI readiness.

Chapter 1 — Strategy & Culture

Whether AI is integrated into organisational strategy, leadership commitment, and cultural readiness for change.

Chapter 2 — Data Foundations

Data quality, governance, accessibility, and whether the data infrastructure can support AI workloads.

Chapter 3 — Governance & Ethics

Oversight structures, accountability mechanisms, bias awareness, and responsible AI policies.

Chapter 4 — People & Skills

AI literacy across the workforce, dedicated capability, and upskilling plans.

Chapter 5 — Execution & Operations

Track record of delivering technology projects, tooling, and operational readiness to run AI systems.

Scoring

The 6 Scoring Dimensions

Each dimension is scored 0–100 based on your specific answers. The overall maturity score is a weighted average of all six dimensions. Dimension scores are constrained by deterministic anchors — they cannot exceed mathematically defined ranges regardless of AI model output.

Dimension	Weight	What it measures
Data	25%	Quality, governance, and accessibility of organisational data for AI use
Governance	20%	Oversight structures, ethical frameworks, and accountability for AI decisions
Strategy	20%	Executive commitment, strategic alignment, and cultural readiness for AI
People	15%	Workforce AI literacy, dedicated capability, and skills development
Execution	15%	Delivery track record, tooling, and operational readiness to run AI systems
Regulatory	5%	Awareness of and preparedness for applicable AI and data regulations

Maturity levels are determined by overall score:

Level	Score range	What it means
Nascent	0–20	AI is not yet part of organisational thinking. Foundational work is needed before any AI projects.
Emerging	21–40	Early awareness exists but no structured approach. Key gaps in data, governance, or accountability remain.
Developing	41–60	Some groundwork is in place. Ready for a focused pilot with proper governance.
Established	61–80	Strong foundations across most dimensions. Ready to scale AI initiatives.
Advanced	81–100	Mature AI capability with governance, measurement, and continuous improvement in place.

Dimension letter grades: A (80–100), B (60–79), C (40–59), D (20–39), F (0–19).

AI models

The Multi-Model Pipeline

No single AI model is good at everything. We use three specialised models in sequence, each doing what it does best. The deterministic scoring engine runs in parallel — AI models cannot override the scores it produces.

1

Company Intelligence

Large language model

Analyses your website to extract your company profile, industry, geography, size, and technology signals. This context shapes everything that follows — the questions adapt and the report is calibrated to your sector.

2

Regulatory Mapping

Live regulatory intelligence

Performs real-time research to map your jurisdiction and industry to applicable regulations including the EU AI Act, GDPR, and sector-specific requirements. Updated at report generation time with current sources.

3

Industry Research

Real-time search with citations

Conducts real-time web research on AI adoption trends, challenges, and opportunities specific to your industry and jurisdiction. Includes live citations from authoritative sources published within the last 12 months.

4

Structured Analysis

Reasoning model

Analyses your 17 assessment responses against deterministic score anchors and consistency checks. Produces dimension scores constrained by mathematical bounds — the model cannot produce scores outside the ranges your answers define.

5

Executive Synthesis

Large language model

Synthesises the analysis into the executive summary, industry context narrative, regulatory exposure section, and strategic recommendations — all tailored to your specific situation, sector, and maturity level.

All models run with structured output constraints and temperature settings tuned for accuracy. Token budgets are enforced to ensure substance over verbosity.

Compliance

The Regulatory Engine

Regulatory mapping is powered by a live intelligence engine that performs real-time research at report generation time. This ensures your report reflects current regulatory developments rather than static rules that may be months out of date.

The engine maps your jurisdiction and industry to applicable regulations, including:

EU AI Act — risk classification, prohibited uses, and compliance timelines for EU-based or EU-facing organisations
GDPR / UK GDPR — automated decision-making, profiling, and data subject rights relevant to AI systems
Sector-specific requirements — financial services (FCA, MiFID II), healthcare (MDR, MHRA), and others where applicable
Emerging frameworks — NIST AI Risk Management Framework, US AI Executive Order provisions

This is not legal advice. Regulatory information is provided for awareness and planning purposes only. Regulations change frequently and their application depends on specific facts. Always consult qualified legal counsel before making compliance decisions.

Limitations

What This Assessment Cannot Tell You

It is self-reported. Scores are based entirely on your answers. If answers are incomplete, optimistic, or inconsistent, the report reflects that. The consistency checker flags obvious contradictions but cannot verify claims.
It is not a technical audit. It assesses organisational readiness, not technical infrastructure. It does not examine your actual data, systems, or code.
It is a point-in-time snapshot. AI readiness changes. We recommend reassessing every 6–12 months, or after significant organisational changes.
Industry benchmarks require sufficient data. Percentile comparisons require a minimum number of assessments from comparable organisations. In sectors with fewer responses, benchmarks are suppressed.
Regulatory information is for awareness only. See the note above. Do not rely on this assessment for legal compliance decisions.
AI models can make errors. The narrative sections are generated by AI and may occasionally contain errors or miss nuances specific to your organisation. The deterministic scoring engine is not subject to this limitation.

Evidence

Evidence Base

The scoring anchors, dimension weights, and maturity descriptors were developed using the following research and frameworks:

McKinsey Global AI Survey (2023–2024) AI adoption rates, organisational barriers, and maturity patterns across industries and geographies.
MIT Sloan Management Review — AI & Business Strategy Framework for assessing strategic AI readiness and the relationship between AI investment and business outcomes.
Gartner AI Maturity Model Maturity level definitions and dimension weighting principles.
NIST AI Risk Management Framework (AI RMF 1.0) Governance and accountability dimension anchors; risk management practices.
EU AI Act (Regulation 2024/1689) Risk classification tiers, prohibited uses, and compliance obligations for high-risk AI systems.
OECD Principles on AI Responsible AI principles used to anchor the governance and people dimensions.
Deloitte State of AI in the Enterprise (2023) People and skills benchmarks; AI literacy measures.
World Economic Forum — AI Governance Alliance Cross-industry governance practices and regulatory readiness benchmarks.
Stanford HAI AI Index Report (2024) Industry adoption data used to calibrate industry context narratives.

Transparency

AI Capabilities

Three specialised AI capabilities work together to produce your report, each applied to the stage it handles best.

Real-time Research

Live web search with citations for industry context and regulatory intelligence. Provides current, cited context for your sector and jurisdiction — updated at every report generation.

Structured Analysis

Reasoning model for structured analysis. Scores your responses, identifies patterns and gaps. Operates within deterministic score bounds that your answers establish.

Synthesis

Large language model for company intelligence extraction and executive synthesis. Produces the narrative sections of your report. Used in both the intake and synthesis stages.

Deterministic components (score anchors, consistency checker) are not AI — they are rule-based systems that produce the same output for the same input, every time.