Skip to content

Resources · Governance tools

Governance tools for practical AI decisions.

Seven calculators and checklists for the questions that come before a rollout: what it is worth, what it costs to govern, who should review it and what to ask a vendor. Each one shows its method and gives you a result without a sign-in.

  • Runs in your browser
  • No sign-in
  • Method and rules shown on every page
Start with the ROI calculator See all seven tools
Connected governance · Example data
ONE DECISIONWhat to govern firstINPUTS STAY ON YOUR DEVICEROI & PAYBACKGOVERNANCE COSTRISK TRIAGEMATURITYVENDOR EVIDENCEAGENT PERMISSIONSREGULATORY SCOPEEXAMPLE DATA · SAMPLE INTERFACE · SEVEN TOOLS, ONE QUESTION EACH
ONE DECISIONWhat to govern first

Illustrative sample graphic. Seven tools, one question each; the records they produce stay on your device.

Size the value

Put numbers on the decision.

  • Calculator

    AI Project ROI & Payback

    Model capacity value and cash savings separately, with payback, break-even adoption and low, base and high scenarios.

    • Capacity vs cash, side by side
    • Payback month
    • Break-even adoption
    • Low · base · high
    CUMULATIVE NET BENEFIT · 12-MONTH SPAN · EXAMPLE VALUESBREAK-EVEN LINEMONTH 0MONTH 6MONTH 12SETUP −$20,000PAYBACK · MONTH 4CAPACITY VIEW · +$58,000 NET · ROI 152.63%CASH VIEW · −$14,000 NET · PAYBACK ≈ MONTH 40 (EST.)CAPACITY · FREED TIME AT LOADED COSTCASH · 25% REALIZATION
    NET BENEFIT · 12 MOCAPACITY +$58,000PAYBACK · MONTH 4CASH −$14,000MONTH 0MONTH 12

    Illustrative preview · example values

  • Calculator

    AI Governance Cost & Savings

    Size the yearly review and evidence workload, the share that can be reduced and the break-even effort reduction.

    • Baseline hours and cost
    • Hours freed vs cash saved
    • Break-even reduction
    • Year rows as CSV
    250 SYSTEMS × 2 REVIEWS A YEAR · 2 h EACH · EXAMPLE VALUES· · ·BASELINE1,000 h · $80,00040% CUTAFTER600 h STAY400 h FREED$32,000 CAPACITYCASH SAVED $0REALIZATION 0%YEAR-1 COST $22,000 · CAPACITY ROI 45.45% · BREAK-EVEN CUT 27.5%
    REVIEW HOURS A YEARBASELINE1,000 h−40%AFTER600 h STAY400 hFREED

    Illustrative preview · example values

Planning estimate from your own assumptions. It is not a forecast, a quote or financial advice, and it does not promise any saving or return.

Review before rollout

Find the questions to settle first.

  • ONE AI USE CASE · TWO LIMITED RULE SETSEU AI ACTSCOPE + CLASSIFICATION REVIEWNYC HIRING TOOLAEDT SCOPE REVIEWOTHER LAWSNOT ASSESSEDFICTIONAL EXAMPLE · NO COMPLIANCE VERDICT
    ONE AI USE CASE · TWO LIMITED RULE SETSEU AI ACTSCOPE + CLASSIFICATION REVIEWNYC HIRING TOOLAEDT SCOPE REVIEWOTHER LAWSNOT ASSESSEDFICTIONAL EXAMPLE · NO COMPLIANCE VERDICT

    Illustrative preview

    Checklist

    AI Regulatory Applicability Screener

    Identify selected EU AI Act and NYC hiring-tool review questions, missing facts and evidence to collect.

    • Scope review topics
    • Open facts
  • REVIEW PRIORITY · ILLUSTRATIVE SCORE · EXAMPLE VALUESSTOP · LEGAL REVIEW FIRSTHIGH · REVIEW BEFORE LAUNCHMEDIUM · REVIEW SOONLOW · STANDARD CHECKSUSE CASE · HIRING SCREENER17 of 26 pointsDECIDES · PERSONAL DATASPOT CHECKS ONLYFLOOR: CONSEQUENTIAL PURPOSE + AI DECIDES → AT LEAST HIGH026LOW 0–6 · MEDIUM 7–13 · HIGH 14–26 · STOP ONLY BY FLOOR
    REVIEW PRIORITYSTOPHIGHMEDIUMLOWHiring screener17 of 26FLOOR → AT LEAST HIGH

    Illustrative preview

    Checklist

    AI Use-Case Risk Triage

    Set a review priority for one AI use case, see the answers that drove it and the reviewers to involve.

    • Review priority + first action
    • Illustrative score of 26
  • SIX GOVERNANCE AREAS · SHARE OF 9 POINTS EACH · EXAMPLE VALUESINVENTORY · 67%ACCOUNTABILITY · 89%POLICY & CONTROLS · 78%EVIDENCE & RECORDS · 22%RISK REVIEW · 56%MONITORING · 33%50% GAP LINEOVERALL 31 / 54 · 57.41%DEFINED ON POINTS ALONECAP → DEVELOPINGWEAKEST: EVIDENCE & RECORDS
    INVENTORYROLESPOLICYEVIDENCERISKMONITORING50% LINE

    Illustrative preview

    Checklist

    AI Governance Maturity

    Rate six governance domains from self-reported answers and get the gaps to close first.

    • Overall level + share of points
    • Six area bars, gaps first
  • 17 ITEMS · 6 EVIDENCE SEEN · 11 OPEN · EXAMPLE VALUESNO TRAINSAYS YESLIMITED USESEENRETENTIONNOT OFFEREDDELETIONSEENSUBPROCSEENSUB NOTICENOT SURESEC REPORTSAYS YESPEN TESTNOT OFFEREDMODEL CHGSAYS YESPINNINGNOT SUREDPASEENAI TERMSSAYS YESINCIDENTSAYS YESTESTINGNOT OFFEREDLIMITSSEENEXPORTSEENEXITNOT SUREDOCUMENTS TO REQUEST · PRIORITY ORDERCRITICALNo training on your dataVENDOR SAYS YESCRITICALIndependent security reportVENDOR SAYS YESCRITICALIncident noticeVENDOR SAYS YESHIGHPenetration test summaryNOT OFFEREDHIGHRetention period and settingsNOT OFFERED+ 4 MEDIUM · 2 LOWNO SCORE · NO VERDICT · ONLY EVIDENCE SEEN CLOSES AN ITEM
    17 ITEMS · 11 OPENSEEN 6SAYS YES 5NOT OFFERED 3NOT SURE 3REQUEST FIRSTCRITICALNo-training clauseCRITICALSecurity reportHIGHPen test summary

    Illustrative preview

    Checklist

    AI Vendor Due Diligence

    Turn what a vendor has shown you into prioritized questions and a list of documents to request.

    • Items by status
    • Questions ranked by priority
  • FOUR OF TEN ACTIONS · WHERE A PERSON COMES IN · EXAMPLE VALUESAGENTOWN LIMITED ACCOUNTREAD TENANT RECORDSDRAFT REPLYSEND TO TENANTSRUN CODE×NNO PERSON · READ-ONLY · NO FLAGAPPROVED EACH TIME · NO FLAGBATCH APPROVAL · HIGHNOT SURE = NO PERSON · CRITICAL8 FLAGS · 1 CRITICAL · 3 HIGH · 3 MEDIUM · 1 LOW
    AGENTRead recordsSend to tenantsRun code×NNO FLAGBATCH · HIGHNOT SURE · CRITICAL

    Illustrative preview

    Checklist

    Agent Permissions & Human Oversight

    Map what an AI agent may do, where a person must approve and what to settle before rollout.

    • Permission map by action
    • Flags by severity

Review aid from self-reported answers. It is not legal advice, a regulatory classification, a certification or an audit, and it cannot see or test any system.

How the tools handle your answers

Your inputs stay on your device.

Everything is worked out in your browser. Anvaylo does not receive the numbers or answers you enter, and nothing is saved when you leave the page. Use estimates and roles, not names, contract text, customer records or credentials.

Worked out here · nothing sent · nothing saved

When you want a second reviewer

Take a result into a human-led review.

Each tool points to the engagement that picks up where the result stops. The record colour shows which tool feeds it.