MIZ OKI SIGNAL — LIVE DEMO PLAN r2.1

Supersedes: # MIZ OKI 3.5/internal/MIZOKI_LIVE_DEMO_PLAN_AUG2026.md (r1.0) Sources: whitepaper r1.2 · r3.5.1 + r3.5.2 · Growth Control r2.0 · completion/activation/OPT-GOV build reports · F4_GO_LIVE_2026-08-24 · PILOT_REPORT_SPEC_v1 · SHOPIFY_INSTALL_ACCEPTANCE_v1. Statuses honest; nothing here upgrades a label.

What changed since r1.0 (why this revision exists)

  1. The SIG-042 passport harness now EXISTS (OPT-GOV Phase 4) — Tier 1's spine is no longer to-be-wired; Phase A becomes "connect harness to the demo surface."
  2. F4 is LIVE — the demo's restraint act can show a real reservation halting at real L2 approval, not a mock. Live beats synthetic; label it as live.
  3. New showable artifacts: dosage curve with CIs · Clipped-ReLU authorization curve (draw the function; the flat zero and the hard clip ARE the pitch) · weekly pilot report (leave-behind) · Shopify 7-step install (Part D) · A/B/C self-test with public pre-registration ("we run our site the way we tell you to run media").
  4. New honesty beat: the deploy gate auto-corrected our own whitepaper (Firestore/Neo4j). Tell it. "Even our docs pass through the gate" is unfakeable credibility.

Demo data provenance (binding for every act)

All figures shown are the SYNTHETIC demo tenant's, watermarked SYNTHETIC on the screen where they appear — including Act 1's ROAS-vs-contribution contrast. Act 2's session trace is the presenter's own consented device session, stated aloud. F4 demo reservations run ONLY against a sandboxed demo tenant with zero-spend caps and demo-only geos — never the pilot tenant, never real spend; if sandbox isolation cannot be proven by config, Act 5 uses the recorded capture, labeled. Demo success metric: demo → pilot-scoping conversations, logged per run.

The 20-minute arc — five acts (Tier 1/2 per r1.0 tiers, unchanged claim ceilings; a 5-minute cut = Acts 1+5 only for exec audiences)

Act 1 — The Lie (3 min). Platform dashboards vs bank account: same campaign, 4× reported ROAS, negative contribution after returns. End on the one-liner: we optimize the number your bank account reports. [L1 Profit-Leak view]

Act 2 — The Anticipation (4 min). Live session trace → VDI/dwell tokens → I-02 forecast with attention summary → hesitation flag → matched-creative rank (logged, not applied — say so). Then the O-1 moment: submit a keystroke-dynamics payload on stage → DISALLOWED_KEYSTROKE_DYNAMICS. Ten seconds; most memorable beat retained from r1.0.

Act 3 — The Proof (4 min). Register a holdout (declared estimand, power) → CATE estimate → refutation battery with one test failing and the estimate WITHHELD. Showing our system catch our own weak estimate outsells any passing metric.

Act 4 — The Profit (4 min). Net-yield waterfall with one row EXCLUDED and the missing cost named → dosage curve: "your profit-maximizing spend is $X, not your budget" (CIs on screen) → J-03 losers list.

Act 5 — The Restraint (4 min). A decision runs the DCP: passport → DEL score plotted ON the Clipped-ReLU curve → below threshold → WITHHOLD with named reason. Then the live F4 reservation halting at approval [LIVE — labeled]. Close: Restraint Ledger [L5] + the Firestore correction story.

Leave-behind (1 min): sample weekly pilot report (ILLUSTRATIVE watermark) + 90-day pilot one-pager.

Build deltas (D-register)

Guardrails (r1.0 §6 retained verbatim + two additions)

All r1.0 guardrails stand: label everything, no fabricated realized outcomes, no hidden happy path, failure is content, ledger discipline. Add: (a) anything shown from shadow/dark surfaces carries its flag state on screen; (b) the anticipation act uses the canon framing — anticipatory intent evidence from content-free signal — never mind-reading language; content_qa gates the demo script itself.

← All docsView source on GitHub →