Frequently asked questions

An independent SaaS platform that scores AI tools used in AEC (architecture, engineering, construction) against a 7-layer compliance framework. It combines tool compliance scores (C1), firm digital maturity assessment (C2), and project context to produce a FitScore — a ranked decision for which tools you can safely deploy on a specific project.

Two audiences: (1) AEC firms — PEs, BIM managers, Innovation Directors, and digital leads who need to verify AI tools before procurement or deployment. (2) AI tool vendors — companies whose tools serve the AEC market and want an independent compliance score to share with enterprise buyers.

No. The ACS score is an independent compliance assessment produced by editorial research — not a legal certification, regulatory approval, or guarantee of compliance. AECO cites and monitors external frameworks (see /standards for the six the methodology grounds to); AECO is not itself a certification body. Always consult your legal and insurance advisors for specific project compliance requirements.

Yes. The platform is fully responsive and accessible on desktop, tablet, and mobile. All core features — maturity assessment, FitScore comparison, registry browsing — are available on any device.

The platform was built by Marcin Kasiak — a practising structural engineer with PhD, PE (Professional Engineer), PMP, and IWE credentials and 20+ years of AEC industry experience. The methodology has been cross-checked across four frontier AI systems (ChatGPT, Claude, Gemini, Grok) for reasoning consistency — not an independent human audit. It is part of AECO.digital — independent intelligence for the AEC technology market.

AECO.digital is a four-app platform with AECO Shield as App 1 (live today). The three coming apps: AECO Bid (pursuit risk + AI exposure + delivery-method fit — see /bid), AECO Strategic (portfolio risk forecast + initiative premortem + AI-investment ROI — see /strategic), and AECO Career (personal licensure + liability heatmap — see /career). Each reads from the same AECO Compliance kernel Shield uses; join the waitlist on any teaser to be notified when it opens.

The AI Compliance Score (ACS) is a 0–10 score produced by the C1 Scanner. It assesses an AI tool across 7 liability layers: Responsible Charge (22%), AI Governance (20%), Code Edition Control (16%), BIM & ISO 19650 (14%), Cybersecurity (12%), Worker Safety (10%), and Vendor Transparency (6%). Scores are evidence-based — derived from public sources, vendor documentation, and editorial research. See /methodology for the full 22-question rubric.

FitScore = ACS × 0.50 + LayerAlignment × 0.30 + MaturityMatch × 0.20. It combines the tool's compliance score with how well its compliance profile matches your project's liability priorities (LayerAlignment) and how well your firm's digital maturity matches the tool's requirements (MaturityMatch). A tool scoring 8.0 ACS may still have a low FitScore if your firm is not ready to govern it safely. See /methodology for the rubric detail.

A three-state classification of a tool's stampability, derived from canonical scoring inputs (ACS + Q21b): Stamp-Supporting (ACS ≥ 8.8 AND Q21b tier ≥ 3 — a Licensed Professional can place responsible charge on outputs), Conditional (AMBER ACS, or GREEN ACS with Q21b tier < 3 — usable with documented conditions), or Not Stamp-Safe (RED ACS or Q21b Tier 1 — output cannot enter stamped deliverables without significant additional governance). This is the first signal shown on every tool page.

A specific assessment question that checks whether a tool's publicly available Terms of Service, EULA, or Acceptable Use Policy contains language that disclaims liability or transfers risk to the user. A score of 2 or below triggers a red warning banner: the tool's legal terms may not support responsible charge on stamped deliverables. This reflects 2026 industry practice where vendor ToS language is increasingly cited in E&O disputes.

Hard filters (HF1–HF5) are binary checks that run before any FitScore is computed. If a tool fails a hard filter, its FitScore is computed and stored but not displayed — the result shows DoNotProcure instead. Hard filters check: jurisdiction licensure (HF1), discipline scope (HF2), E&O insurance coverage (HF3), GL coverage (HF4), and DBE requirements for public projects (HF5).

No hands-on testing. All ACS scores are produced from publicly available evidence: vendor documentation, Terms of Service, published case studies, third-party review platforms (G2, Capterra), regulatory filings, and AEC industry knowledge. Every score includes a methodology disclosure. This approach maintains editorial independence — vendor relationships and payments never influence scores. See /vetting-lab for the full 7-step evaluation pipeline.

Six recognized frameworks: ISO 42001 (AI management systems), NIST AI RMF (US risk-management framework), EU AI Act (regulation classifying AI systems by risk), CMMC 2.0 (US DoD cybersecurity maturity), ISO 19650 (BIM information management), and IBM AI Principles. AECO cites and monitors these — Plan A auto-fetches the open sources, Plan B schedules editorial verification for the bot-blocked ones (ISO / federal WAF). Every detected change routes through an editorial review before methodology citations update. See /standards for the live citation count per framework + monitoring detail.

The public Registry (/registry) lists every published assessment with its ACS score, DCS verdict, and Stamp-Safe classification. The Vetting Lab (/vetting-lab) describes the evaluation pipeline itself and shows the currently-published assessments alongside a sign-up for notification when new ones publish. Every published assessment traces back to the same 7-step pipeline.

A 40-question self-assessment that scores your firm across 7 dimensions: Data Maturity (D1), BIM & Digital Infrastructure (D2), AI Readiness (D3), Process & Governance (D4), Talent & Capability (D5), Adoption & Scaling (D6), and Supply Chain Digital Alignment (D7). It also assesses your ISO 19650 adoption stage (D2b). The output is a Firm Maturity Profile (FMP) — used by C3 to personalise FitScore rankings.

Approximately 10–15 minutes for a first assessment. Subsequent reassessments are faster as previous answers are retained for reference.

Your data is yours. Assessment history stays in your account for as long as you maintain access, and every assessment is exportable at any time. When paid subscription tiers ship, retention policy for lapsed subscriptions will be documented here.

If your firm does not have a named person responsible for reviewing AI-assisted outputs before a licensed-professional stamp is applied (D4.5 = No), your Process & Governance dimension score is hard-capped at 4.9 (Developing band) regardless of other answers. This reflects the NSPE responsible charge requirement that a licensed professional must independently verify AI outputs before stamping.

Start with a Self-Audit ($149) — automated ACS scoring with gap report, no badge. Upgrade to Verified Certification ($1,600) for expert human review, official badge, registry listing, and 12-month validity. For federal work, the Federal Pack ($4,500+) covers multi-standard audit with FAR/DFARS documentation. Contact us at hello@aeco.digital to begin.

Target turnaround at commercial launch — Self-Audit: automated, within 24 hours of questionnaire completion; Verified Certification: 5–10 business days including expert review; Federal Pack: 10–15 business days. Editorial evaluations currently run at a slower pace (2 published so far — see /vetting-lab); commercial-tier SLAs commit at commercial launch.

Yes — the US Market Entry Pack ($3,500–$6,000) is specifically designed for international vendors. It maps your existing certifications (ISO 42001, EU AI Act compliance, CE marking) to US requirements (NIST AI RMF, CMMC 2.0, SOC 2, state AI laws) and produces a gap analysis with a prioritised remediation roadmap.

No. Payment covers the assessment process — not the outcome. A tool that fails the 7-layer framework will receive a low score regardless of which tier was purchased. Editorial independence is a foundational principle. Vendor relationships and payments never influence published scores.

The public Registry — browse every listed tool, view ACS scores and score bands, and filter by category or jurisdiction. Running your own tool assessments, comparing tools against your firm profile, generating PDF reports, and firm-wide features all require a paid Pro ($149/user/mo) or Firm ($799/mo) subscription. Both include a 14-day free trial — card required at signup, no charge if you cancel during the trial.

Pro includes Tool Check (C1 ACS scoring), FitScore Compare (up to 8 tools per batch), Project AI Sign-Off (C4), Public Registry access, methodology-stamped audit artifacts, core PDF reports, and 7-year audit retention (per our published Quality Policy). See /pricing for the current tier list.

Yes — Firm plans include shared audit trail across the team, firm-wide access + shared data scope, the Firm Maturity Profile (C2), and all 9 procurement-ready PDF report generators. See /pricing for the current tier list. Enterprise engagements add dedicated success manager + SLA, with SSO/SAML, API access, and multi-region data residency on the roadmap.

All data is stored in Supabase (PostgreSQL) with row-level security. The platform is deployed on Vercel with global CDN. Your firm's assessment data is private by default — only visible to authenticated users in your account. Public registry data is visible to all users.

The platform uses Supabase authentication with row-level security policies, HTTPS everywhere, and no payment card data on our servers. Payments are processed by Stripe; card data does not pass through AECO systems. Environment variables and API keys are never exposed client-side.

The Anthropic Claude API (claude-sonnet) powers narrative generation for FitScore reports, gap analysis summaries, and coaching recommendations. The API is called server-side only — your assessment data is not used to train AI models.