v1.24.1 · Responsive subtitles

Subtitles use the available blue caption bar instead of preserving the subtitle file’s fixed line breaks. Short cues stay on one line when they fit; narrow screens wrap naturally. Slide players and in-page explanations share this behavior. Audio, subtitle timing, and the v1.24.0 PDF editions remain unchanged.

v1.24.0 · Presentation and decision clarity

Larger, stable slides; on-demand audio; outcome-first routes with matching PDFs; explicit pilot scope and ownership; separate validation and release publication.

Current downloads are below. Superseded editions remain in immutable GitHub releases. Send a guided PDF for a short meeting and the full deck for reference.

Prepared by Tom Wu · Contact / feedback. AI-generated illustrations and synthetic English narration. Diagrams and scenario models are authored explanations; sources retain their own attribution. Narration provenance.

Publication editions

The version identifies the site and presentation player. A player-only patch can retain the previous slide-content and PDF edition, as noted below. Source publication dates remain separate from the site release date. All diagrams are proposed designs or authored synthesis unless explicitly described as study results.

GitHub releases & assets ↗Open the presentation room →

v1.23.0 · 9 September 2026

Presenter controls preserve Auto-next across pause, resume, replay and manual navigation. The environment lifecycle follows the recording’s measured caption cues. Choosing a failure, Next step or Overview pauses the explanation; resuming restores its narrated baseline.

Pilot decisions now follow the sponsor-selected outcome: effort, cash, quality, reliability, evidence or reuse. Every route retains quality floors, evidence requirements and a funded cost ceiling. Cash claims require Finance-validated capture. Ranking-only discovery answers leave activity shares unknown.

The existing banking comparison now accepts three evidence-linked observations: existing QE, modernized QE with AI off, and AI on with the same foundation. Modernization and incremental AI effort are calculated separately, including slowdowns. Missing or unmatched controls remain unknown. Shared-platform readiness asks which workflow is being reused and distinguishes second-team validation from supported rollout.

A new narration review gate binds explanations to rendered content; semantic changes require an explicit review or a refreshed recording. Release identity and package integrity are checked before Pages deployment, including existing tag targets. All four PDF editions advance to v1.23.0, with 116 narrated slides and an updated pilot-decision recording. GitHub release v1.23.0 includes the current publication and narration review record.

v1.22.0 · 9 September 2026

The banking case now follows nine explicit AI-assisted steps, from approved requirements and developer unit tests through API/UI tests, fixtures, CI execution, investigation, repair and release evidence. A connected artifact trail preserves the original injected-defect failure, reviewed fix and fresh retest.

The comparison distinguishes manual QE, modernized QE and AI-assisted QE, with separate contributions and measures. Existing slide links remain valid. Guided banking routes bring the new workflow, comparison and evidence trail into the presentation sequence; the industry decks link to the banking implementation story.

All 116 slides have narration and timed English captions. Changed explanations use new recordings; workflow, comparison and artifact highlights follow those recordings, including pause and seek. Existing two-second slide pauses and shared audio focus remain. All four slide/PDF editions advance to v1.22.0: banking 21 executive / 33 technical, industry 26 executive / 36 technical. The film retains its original edition. GitHub release v1.22.0 packages the current publication.

v1.21.2 · 9 September 2026

The 3D architecture now opens from a higher, more front-facing camera angle, so the platform’s depth is easier to read. Framing fits the model with space around its edges and preserves its proportions across wide, short and mobile windows. Camera motion keeps the same elevated view.

Reset view restores both the original angle and zoom, including after an orbit drag. The model, recordings, narration synchronization and original downloadable film retain their existing editions. GitHub release v1.21.2 packages the current publication.

v1.21.1 · 9 September 2026

Starting audio now pauses other AI × QE players across open tabs and windows on the same site, page explanations and nested embedded presentations. Playback requests claim focus before loading finishes; a cancelled request cannot interrupt the newer player when its audio arrives.

Taking over audio also cancels the previous deck’s pending automatic transition. The existing two-second pause between slides, synchronized captions and diagram highlights remain. Interrupted players stay paused until selected again. Refresh existing presentation tabs to load this player update.

Recordings, scripts and slide/PDF editions are unchanged. GitHub release v1.21.1 packages the current publication.

v1.21.0 · 9 September 2026

The technical platform slide and its existing page explanations now have a 4-minute-50-second full walkthrough, replacing the short technical summary. The recording explains all 11 components across 13 synchronized sections, with 72 audio-timed English captions.

An authored payment-retry testing scenario connects approved expected behavior, versioned context, AI drafting, authorized execution, isolated tests, release review and retained evidence. The walkthrough also explains the separate AI application evaluation path, reviewed regression cases and QE modernization prerequisites. The executive summary retains its shorter recording.

Replay clears the completed status when an explanation starts again. Existing pages, slide/PDF editions and the four 3D scenario recordings are retained. GitHub release v1.21.0 includes the new technical recording, transcript, captions and visual cues in the narration bundle.

v1.20.0 · 8 September 2026

The 3D architecture demonstration now follows its English audio across all four scenarios. Select Listen to explanation: the spoken section, highlighted components, flowing arrows and captions share the recording’s timeline. Both sets of playback controls pause, replay, seek and change speed together.

The 24 authored audio sections use exact caption anchors. A section can highlight several components when the recording discusses them together. The denied-action path never invokes a tool, and the failed-proof scenario stays on hold when the narration describes an alternative passing path. Reduced motion retains static highlights. Explore without audio returns to the original stage walkthrough.

Existing recordings, captions, PDF editions and the original 49-second silent film are retained. The GitHub release v1.20.0 includes the 3D cue definitions in the narration bundle.

v1.19.0 · 8 September 2026

Active destinations now use a gold fill, strong border and readable dark text across all nine animated 2D workflow diagrams. Highlights remain visible when paused and in the mobile readable overview. Other components and connections remain legible.

The existing page explanations and all four decks now use 78 caption-anchored visual cues across 12 recordings to synchronize these diagrams with the spoken explanation. Audio controls the highlighted components and arrow position; pause, seeking, speed changes and replay stay on the same timeline. Manual flow controls pause narration and return control to the presenter. Rejection and quarantine stay separate from acceptance and repair.

The technical platform slide and the homepage architecture demonstrate the update. The existing recordings, captions, two-second slide pause and PDF editions are retained. The 3D demo still uses its separately labelled scenario overview. GitHub release v1.19.0 includes the new visual cue definitions in the narration bundle.

v1.18.1 · 8 September 2026

The audio narration panel now matches the slide width with a 12-pixel gap and consistent borders across standalone and embedded presentations. It stays next to its slide on tall screens, with separate room for diagram controls and navigation.

Voice-name branding has been replaced with Audio narration in diagram players, transcripts and publication copy. The existing 109 recordings, synchronized captions, two-second slide pause and PDF editions are retained. GitHub release v1.18.1 includes the narration download.

v1.18.0 · 8 September 2026

Added a two-second breathing pause after each narrated slide across all four decks. Pause holds the transition; manual navigation, opening notes, hiding the tab or turning Auto-next off cancels it. The pause stays two seconds at every playback speed.

All 109 slides now include their spoken explanation in the existing Sources & notes drawer. Existing diagrams, readiness and modernization sections, banking visuals, research charts and all four 3D architecture scenarios offer Audio explanations, synchronized English captions and expandable narrator notes in place. Playing one explanation pauses other narration, including embedded presentations. Calculator recordings explain the published baseline; 3D recordings explain the scenario overview separately from its stage timer.

No new pages or audio recordings were created. Narration retains v1.17.0; banking PDFs remain v1.16.1, and industry PDFs remain v1.12.2. GitHub release v1.18.0 packages the current assets.

v1.17.0 · 8 September 2026

Added English audio narration to all 109 slides across the four presentations. Select Play narration to listen from the current slide with synchronized captions and automatic advancement when each recording finishes. Pause, replay, seek, playback speed, a caption toggle and a readable transcript support both full decks and guided stories.

Captions use the recordings’ character timestamps, with technical names restored for reading. The scripts explain the diagrams and distinguish the proposed banking scenario from reported industry evidence. Narration assets use edition v1.17.0; banking PDFs remain v1.16.1, and industry PDFs remain v1.12.2.

GitHub release v1.17.0 includes the current PDFs and source registers, plus a downloadable narration bundle containing slide audio, subtitles and transcripts.

v1.16.1 · 8 September 2026

Clarified the banking before/after story: 75 offshore QA staff include 45 manual/domain testers, with few shared test environments, no backend service virtualization and limited vendor test capacity. Manual setup, reruns and evidence collection compete for those environments; adding staff alone does not increase available vendor slots.

The executive and technical architecture comparisons, workflow baseline, application landscape, readiness assumptions and modernization backlog now reflect these constraints. Virtualized service tests and isolated runs remain proposed capabilities; real-provider and settlement checks still need reserved capacity. No vendor environment count or measured improvement is invented.

Banking slides/PDFs advance to v1.16.1, retaining 19 executive / 30 technical slides and existing anchors. Industry PDFs remain v1.12.2. GitHub release v1.16.1 packages the current assets.

v1.16.0 · 8 September 2026

Added before/after architecture and outcomes beyond hours. The technical comparison shows workstations, runners, application/database boundaries, provider models and retained evidence. The executive comparison shows how squads move from individual tools to supported shared QE services.

Three primary pilot measures cover payment-fault detection, critical-scenario coverage and complete release-decision evidence; repeatability, real-integration fidelity and second-squad reuse supply supporting proof. Definitions include denominators, proposed acceptance criteria, owners and limits. Baseline and observed after results remain not recorded. The measurement approach separates QE modernization from the incremental contribution of AI.

Banking slides/PDFs advance to v1.16.0: 19 executive / 30 technical slides. Existing anchors remain valid; guided routes and featured jumps include the new material. Industry PDFs remain v1.12.2, questionnaire v4, research companion v1.7.0 and film v1.8.0. GitHub release v1.16.0 packages the current assets.

v1.15.0 · 8 September 2026

Rebuilt the banking technical briefing as an engineering walkthrough: separate authoring and execution boundaries, a payment API contract, timeout/retry sequence, versioned context and fixtures, deterministic monetary assertions, Azure CI publication, failure routing, evidence schemas and correlation IDs. Readiness now requires executable proof; adapter contracts define inputs, outputs and failure behavior. The guided technical route follows those engineering decisions.

The executive briefing retains its strategic-vision narrative. Both banking PDF editions advance to v1.15.0 for a consistent release pair, retaining 17 executive / 28 technical slides and stable slide anchors. Product task syntax is source-backed; the reference architecture and code examples are labelled authored proposals or pseudocode. Industry PDFs remain v1.12.2, questionnaire v4, research companion v1.7.0 and film v1.8.0. GitHub release v1.15.0 packages the current assets.

v1.14.1 · 8 September 2026

Renamed the current client scenario to Our Banking Client across the case explorer, adoption roadmap, readiness and modernization material, diagrams, presentation titles, source notes and downloadable trial briefs. The illustrative scenario disclosure and existing links remain valid.

The fintech slide/PDF editions advance to v1.14.1, retaining 17 executive / 28 technical slides. Industry PDFs remain v1.12.2, questionnaire v4, research companion v1.7.0 and film v1.8.0. Earlier release entries and archived editions retain their original names. GitHub release v1.14.1 packages the current assets.

v1.14.0 · 8 September 2026

Added the AI adoption roadmap: individual copilots, connected workflow assistants, bounded QE agents and a shared QE AI platform. The interactive PAY-142 examples connect inputs, AI authority, outputs, reviewers, technology, dependencies and evidence needed to expand. Platform foundations start on day one; teams can scale useful assistance without delegating more actions.

Executive slide 17 explains the four layers; technical slide 28 maps their integration paths. The fintech case, primary menu, readiness page and modernization page link to the roadmap. Primary GitHub and Microsoft documentation supports product capabilities; the roadmap and client examples are authored proposals, without fixed timing or savings claims.

Fintech slides and PDFs advance to v1.14.0: 17 executive / 28 technical slides. Industry decks remain v1.12.2, questionnaire v4, research companion v1.7.0 and film v1.8.0. GitHub release v1.14.0 packages the current assets.

v1.13.0 · 8 September 2026

Made the executive before/after slide explicit and added a technical API comparison showing manual preparation, shared execution and fragmented evidence alongside reviewed AI drafts, repeatable tests and linked proof. The after state is a pilot target with named platform prerequisites. Both views also appear in the case explorer.

Renamed the director conversation to Brainstorm questions. Reframed the engagement slide as Observed time framework for the API pilot, with baseline, repeated API runs and second-team reuse phases. Example windows are labelled planning inputs; actual observations remain not recorded. The implementation notes include a record for dates, effort, waits, foundation work and quality.

Fintech slide/PDF editions advance to v1.13.0: 16 executive / 27 technical slides. Existing anchors remain valid; guided routes include the before/after comparison and pilot timing discussion. Industry decks remain v1.12.2, questionnaire v4, research companion v1.7.0 and film v1.8.0. GitHub release v1.13.0 packages the current assets.

v1.12.3 · 8 September 2026

Labelled the timeout-and-retry example explicitly as a QE testing scenario in the case explorer, executive and technical slides, payment diagram and implementation notes. The explanation identifies a deliberately injected duplicate-posting defect and the expected one-transfer outcome. A timeout alone does not imply two debits; the example is not a reported client incident.

The two fintech slide/PDF editions advance to v1.12.3 (16 executive and 26 technical slides). Industry decks remain v1.12.2, the questionnaire v4, research companion v1.7.0 and film v1.8.0. Slide links and test behavior are preserved. GitHub release v1.12.3 packages the current assets.

v1.12.2 · 8 September 2026

Standardized the audience label to Executive and the scenario name to FinTech Client across current pages, diagrams, presentations, search results and downloadable briefs. Simplified scenario wording while retaining the disclosure that staffing, systems, hours and targets are illustrative planning inputs requiring client validation.

All four slide/PDF editions advance to v1.12.2, with 25 / 35 / 16 / 26 slides (102 total). The fillable discovery questionnaire advances to v4 with the updated audience wording. The research companion remains v1.7.0 and film v1.8.0. Existing presentation and readiness links remain valid. GitHub release v1.12.2 packages the current assets; earlier editions below are historical records.

v1.12.1 · 8 September 2026

Renamed the fictional client from Harbor to Northstar Payments, a digital payments provider, across the fintech case, implementation plan, readiness assumptions, architecture diagrams, audience decks and downloadable trial briefs. The payment retry scenario and its illustrative staffing, effort and adoption assumptions are unchanged. Existing case and readiness links still work.

All four slide/PDF editions advance to v1.12.1, including the shared payment architecture caption in the industry technical deck. Counts remain 25 / 35 / 16 / 26 slides (102 total). Historical releases retain their original names and editions. The research companion remains v1.7.0, questionnaire v3 and film v1.8.0. GitHub release v1.12.1 packages the current assets.

v1.12.0 · 8 September 2026

Connected the modernization-to-AI story with an interactive dependency matrix, concrete Harbor tool integration architecture, before/after operating model, offshore handoff swimlane and eight PAY-142 sample artifacts. The environment lifecycle now demonstrates passing checks, failed assertions and setup failures with retained evidence and cleanup. Selected bank cases have compact intervention/outcome stories; AI, human review, virtual services, systems and failed checks have distinct visual roles.

Diagnosis-only pilots now use their own readiness scope, separate from execution and retesting. Guided fintech stories connect vision, evidence, architecture, dependencies and discussion while preserving original slide anchors. Full decks remain available in their established order.

All four slide/PDF editions advance to v1.12.0: industry Executive 25 slides, industry technical 35, fintech Executive 16, fintech technical 26 (102 total). Two technical views extend the existing deck; several text-heavy slides become visual explanations. The research companion remains v1.7.0, questionnaire v3 and film v1.8.0. GitHub release v1.12.0 packages the current publication assets.

v1.11.1 · 7 September 2026

Made QE modernization an explicit fintech adoption dependency. Six capability areas now connect to proof and accountable roles. Each client trial and downloaded brief identifies its own modernization prerequisites and opens the matching readiness assessment with unknown client assumptions. Required gaps need funded remediation before broader AI execution; reviewed drafting follows its own prerequisites.

This site and trial-brief clarification retains the four v1.11.0 slide/PDF editions and their 100 slides. The research companion remains v1.7.0, questionnaire v3 and film v1.8.0. GitHub release v1.11.1 includes the same 15 publication assets.

v1.11.0 · 7 September 2026

Added Fintech results and client pilots: seven cases, ten sources, reported-outcome charts, six adoption practices and a downloadable trial discussion brief for onboarding design, payment API tests or triage. The architecture walkthrough shows a rejected candidate stopping before the maintained suite. Cases retain their evidence type, measurement scope and limitations.

All four slide/PDF editions advance to v1.11.0: industry Executive 25 slides, industry technical 35, fintech Executive 16, fintech technical 24. Eight new slides preserve prior anchors. The research companion remains v1.7.0, questionnaire v3, and film v1.8.0. GitHub release v1.11.0 includes 15 assets: six PDFs, the film and captions, three CSV/JSON source registers, and checksums.

v1.10.0 · 7 September 2026

Added QE modernization to the conversation path. Six workstreams connect reliable tests, containerization, virtual services, data, CI evidence and shared ownership to broader AI adoption. A guided test-environment architecture and API, web/mobile and legacy examples distinguish real software, simulated dependencies and separate integration checks. Thirteen primary-source records explain the evidence and limits.

The Harbor implementation now includes a concrete modernization sequence and explicit foundation costs. Reviewed assistance can begin alongside this work; broader execution depends on evidence for the selected scope.

All four slide/PDF editions advance to v1.10.0: industry Executive 23 slides, industry technical 33, fintech Executive 14, fintech technical 22. Existing slide anchors remain stable. The research companion stays v1.7.0, questionnaire v3, and architecture film v1.8.0. GitHub release v1.10.0 includes the current PDFs and both source registers.

v1.9.1 · 7 September 2026

The desktop navigation now stays at 248 px instead of expanding with the browser window. The recovered space goes to the page, with a centered content area capped at 1,200 px. Navigation labels and click targets retain their size; phones retain the collapsible Menu.

This is a layout patch. All four slide/PDF editions remain v1.9.0, the research companion v1.7.0, questionnaire v3, and architecture film v1.8.0. GitHub release v1.9.1 includes the current downloads.

v1.9.0 · 7 September 2026

Added Platform readiness to the primary menu and discovery path. Twelve adoption assumptions connect infrastructure, DevOps, service virtualization, data, testability, frameworks, evidence, AI access, people, platform ownership and funding to each workflow. The interactive worksheet identifies required capabilities and unresolved evidence without averaging away blockers.

The fintech case and implementation register now make the unverified foundations explicit. Pilot and economics guidance require a funded remediation backlog before committing to scope, timing or benefits. New editable diagrams explain the adoption foundations and the boundary between virtualized dependencies and real integration.

All four slide/PDF editions advance to v1.9.0: industry Executive 22 slides, industry technical 31, fintech Executive 13, fintech technical 20. Existing slide anchors are preserved. The research companion remains v1.7.0, questionnaire v3, and architecture film v1.8.0. GitHub release v1.9.0 contains the current downloads.

v1.8.1 · 7 September 2026

Fixed search results disappearing before their links could open in Safari. Results remain available through the click, with native navigation to pages, dictionary terms and individual slides. Search scripts now use the site version so returning readers receive the fix. Automated checks exercise actual clicks, taps, keyboard selection and opening results in a new tab in Chromium and WebKit.

GitHub release v1.8.1 includes the current publication assets. Slide/PDF and research editions retain v1.7.0, the questionnaire remains v3, and the architecture film retains v1.8.0.

v1.8.0 · 7 September 2026

A new visual identity and conversation-led navigation connect the strategic vision, fintech case, architecture demonstration and discovery guide. The presentation room brings all four audience decks together with audience filters, outlines and direct presentation and PDF links. Research, economics and delivery controls remain available as supporting evidence.

GitHub release v1.8.0 includes the current PDFs, architecture film, captions, source registers and SHA-256 checksums. Future versions publish a GitHub release after the site deploys successfully.

This is a site and presentation-player update. All four slide/PDF editions and the research companion retain v1.7.0; the fillable questionnaire remains v3. Slide content is unchanged. The architecture film is refreshed to display site v1.8.0.

Banking scenario · Executives & sponsors
21 slides · PDF v1.24.0 ↓Banking scenario · Technical SDs & leads
33 slides · PDF v1.24.0 ↓Industry perspective · Executives & sponsors
26 slides · PDF v1.24.0 ↓Industry perspective · Technical SDs & leads
36 slides · PDF v1.24.0 ↓QA research companion
13 pages · PDF v1.7.0 ↓Discovery questionnaire
Fillable PDF v4 ↓

v1.7.0 · 7 September 2026

Focused the publication on AI-assisted quality engineering across the homepage, shared branding, research pages, presentations, diagrams, dictionary and questionnaire. The interactive capability map now includes test maintenance. The technical deck closes with repair, retest and a time-limited flaky-test quarantine, with separate animated paths and an explicit unresolved-work record.

The current library contains 30 sources and the dictionary contains 62 terms. Superseded mixed-scope downloads are no longer published. The architecture film has been refreshed to match the site edition. Fintech case decks and PDFs are also re-exported as v1.7.0 to remove the old branding from PDF metadata; their QA narrative is unchanged. Fintech Executive PDF · Fintech technical PDF.

v1.6.1 · 7 September 2026

Corrected the fintech workflow explorer: field labels now align above their descriptions, and all eight stage buttons form an even grid with four columns on desktop and two on smaller screens. Narrow screens show workflow details in one column. The staffing labels use the same left alignment.

The payment simulation’s stage strip stacks vertically on the narrowest phones so the page stays within the screen. Its arrows continue to follow the flow direction.

Fintech slides and PDFs retain edition v1.6.0, research decks retain v1.3.0, and the architecture film retains its v1.6.0 edition.

v1.6.0 · 7 September 2026

Added the Harbor fintech case explorer: a fictional organization with 75 offshore QA staff and a retry-safe payment release. It includes a step-by-step payment simulation, a shared QA architecture, all eight workflow stages, three delivery-maturity profiles, capacity arithmetic and an implementation plan. Every staffing figure, effort estimate and case outcome is labeled as assumed or illustrative.

The case has two dedicated, QA-focused briefings with diagrams and editable web content:

The industry research decks retain slide edition v1.3.0. Refreshed the existing architecture film’s site edition label to v1.6.0; its scenario content is unchanged.

Fixed the theme’s search focus handling when focus leaves the page or returns to browser controls. A missing next focus target now closes search without a JavaScript error.

v1.5.0 · 6 September 2026

Added the AI × QE dictionary with plain-language definitions, acronyms, concrete examples and related terms across six topics. Search, topic filters and A–Z navigation support quick lookup; each term has a shareable link and appears in the site-wide search.

All 11 architecture components link between the dictionary and the 3D inspector. The homepage, site navigation and briefing notes also link to the dictionary. Definitions remain readable without JavaScript. Refreshed the architecture film’s edition label. Slide content and PDFs remain edition v1.3.0.

v1.4.2 · 6 September 2026

Retains the clearer 3D architecture introduced in v1.4.1. Playback now also checks the current reduced-motion preference on the next visible frame, covering browsers that delay the preference-change event. Switching reduced motion on pauses playback and disables camera motion; switching it off restores the controls without starting playback automatically.

v1.4.1 · 6 September 2026

Made Architecture in motion easier to follow. Inputs, AI generation, checks and evidence now have distinct colors and text badges. Larger labels, a numbered current-step marker and white directional flows identify what is happening. Full map reveals the other connections.

A light explanation panel uses simpler language and a short route summary. The camera stays still by default, each stage lasts seven seconds, and a pace selector supports slower viewing. Updated the 49-second film, captions, poster and editable Blender scene to match. Slide content and PDFs remain edition v1.3.0.

v1.4.0 · 6 September 2026

Added Architecture in motion, a Blender-authored 3D model with a Three.js interactive viewer. Explore 11 modules, four scenarios, moving signals, camera orbit and component inspection. Denied requests stop at authorization; failed checks reach a held release decision. The AI-application evaluation branch remains distinct from generated-test execution.

Includes a downloadable 35-second architecture film, captions, a static fallback and the editable Blender scene. Motion respects reduced-motion settings. Slide content and PDFs remain edition v1.3.0.

v1.3.1 · 6 September 2026

Animated slides now have a visible Play/Pause, Next step and Overview bar below the canvas. It stays clear of the diagram and navigation, including in the embedded player. A guided flow plays once when its diagram enters view; Pause, Overview, motion-off and reduced-motion preferences retain control. Diagram controls provides view options and branch selection.

The slide content and downloadable PDFs remain edition v1.3.0. The page banner and player header identify this playback update as v1.3.1.

v1.3.0 · 6 September 2026

Shared slide frame, consistent margins and typography, persistent navigation, explicit presentation modes, source/edition drawers, mobile diagram overviews and searchable slide anchors. The embedded player preserves the current audience and slide in its share URL.

Added strategic choices and a financial-services workflow to the 21-slide Executive deck; concrete payment reference contracts, failure semantics and a branched evidence workflow to the 29-slide technical deck. Expanded the library to 32 entries and added research coverage and deployment comparison criteria. Reconciled scenario arithmetic, introduced negative task savings, corrected the TestGen denominator and specified the pilot decision bands.

The PDFs are static exports of this slide edition. Animation, scenario controls, component inspection and expanded research notes are available in the interactive decks and linked research pages.

v1.2.1 · 6 September 2026

Clarified the architecture tour with uniform context paths, explicit directed steps, a legend, reset and synchronized pause/resume.

v1.2.0 · 6 September 2026

Expanded the audience presentations to 18 Executive and 26 technical slides, with 17 additional native diagrams covering strategy, test validity, evaluation and operations.

Original research companion · 5 September 2026

The original concise companion had 13 pages and 30 sources. It has been superseded by the QA research companion linked in the current edition above. The document library is the current source register.