acdl/docs/presentations/nova-no-humans-platform-talking-points.md

Nova — The No-Humans Infrastructure Platform: Talking Points

Step 4 of the 4-step deck process. Presenter cues distilled from the source of truth (nova-no-humans-platform.md). 3-6 bullets per slide

• key takeaway. Indexed by Marp slide #. v1.17 — REQ-196, REQ-197

---

Slide 1 — Arc Preview

• Open with the stake line: "18 capabilities verified, 0 consumer estates in production"
• Preview the 5-act arc so the audience knows the structure
• Set the honesty frame: "this is an evidence deck, not a hype deck"
• Key takeaway: you'll leave knowing what's proven, what's pipeline-ready, and what's deferred

Slide 2 — The No-Humans Imperative

• The operator is the bottleneck: days vs. minutes for provisioning
• Key reframing: "no-humans" = no human in normal operations; stage-gate attestation is human by design
• Cite the no-humans thesis doc
• Key takeaway: autonomy in operations, human at stage gates

Slide 3 — Nova's Vision

• Read the vision statement verbatim — it's precise
• Emphasize "provable, not promised" — the difference between marketing and defensible
• State the attestation model up front to prevent mishearing
• Key takeaway: invisible operations with provable trust

Slide 4 — Strategic Objectives + Anti-Goals

• The 4 objectives are the "what"; the 5 anti-goals are the "what NOT"
• Anti-goal #3 (not removing humans from accountability) reinforces slide 3
• Anti-goal #5 (not sold to operators) explains why this deck is for leadership
• Key takeaway: purpose-built for infra ops, sold to leadership on outcomes

Slide 5 — 12–18 Month Targets

• The three-section split (current / post-pilot / deferred) IS the honesty model
• "Partial" means the pipeline works but the denominator is zero (0 consumers)
• The Post-Pilot targets are committed; the numbers fill when a pilot runs
• Key takeaway: which numbers are real today vs. deferred honestly

Slide 6 — The Platform Pipeline

• Walk the pipeline left-to-right: contract → resolver → adapter → plan → policy → confidence → gate → apply
• Key insight: dev is autonomous; qa/prod/dr require attestation
• The confidence signal is the "AI" — 6-input weighted score, not an LLM
• Key takeaway: the path from intent to evidence, with humans at stage gates only

Slide 7 — The Decision Ledger

• The D-122 honesty sentence is critical: "Nova's AI is the confidence-gated policy engine, not an LLM"
• The ledger is the moat: features can be copied, an immutable decision history cannot
• Every decision has outcome backfill from apply.completed
• Key takeaway: autonomous is defensible because every decision is immutable, queryable, accountable

Slide 8 — The 8-Concern Attestation Matrix

• The matrix is not a rubber stamp — it's structured, freshness-validated, SoD-enforced
• Offline-testable concerns run for real; operator-supplied concerns accept signed evidence
• SoD on prod: the approver can't be the same person who built it
• Key takeaway: autonomy in operations, human in accountability, by design

Slide 9 — Telemetry Architecture

• Deliberately minimal (Nova-native, no Kafka/Prometheus/ClickHouse)
• Every number in the Proof act is traceable to a file path
• The hot path is deferred (D-126) — cold store is sufficient for batch
• Key takeaway: the architecture IS the trust substrate — "where does this number come from?" → file path

Slide 10 — Capability Health

• 18V+4S is the single most important proof point
• The 4 Skipped are live-AWS caps — honestly skipped (D-096), not broken
• When live AWS is re-provisioned, they reactivate
• Key takeaway: the platform works, and we're honest about what we can't test

Slide 11 — Decision Ledger + Attestation Coverage

• Both 100% — no AI decision is ever lost; no prod/dr promotion lands without a human sign-off
• The trust snapshot has a chain-integrity verdict (the ledger hasn't been tampered with)
• D-083 (S3 Object Lock + JWS) is the next step for the ledger
• Key takeaway: trust is provable — not a marketing claim, a queryable record

Slide 12 — Zero-Touch Efficiency

• The Post-Pilot caveat is the honesty model: pipeline works, denominator is zero
• This is NOT a fabricated "99% touchless" claim
• The numbers fill when a pilot runs
• Key takeaway: the measurement works; the numbers activate with a pilot

Slide 13 — Cost & ROI

• The ROI formula is shown inline — not hidden in a footnote
• The N=0 caveat is stated explicitly
• This is the "no fabrication" constraint in action
• Key takeaway: the formula is ready; the production denominator activates with a pilot

Slide 14 — What's Deferred — and Why

• The preempt is critical: deferrals are measurement infrastructure, not autonomy
• The platform IS autonomous in operations; what's deferred is the evidence pipeline
• Showing this to leadership demonstrates honesty, not weakness
• Key takeaway: the autonomy is real; the measurement gaps are documented

Slide 15 — Roadmap to the North Star

• Every deferred metric has a specific unblock requirement and a candidate milestone
• The re-evaluation triggers ensure the metrics layer evolves
• Nothing is hand-waved; everything has a plan
• Key takeaway: the path from "honestly deferred" to "here's how we get there"

Slide 16 — Recap + Ask

• Recap the 5-act arc so the audience leaves with the structure
• The ask is a business decision: approve a pilot + the ledger build-out
• "Pipeline-ready" → "production-proven" is the value proposition
• Key takeaway: approve a pilot + the ledger build-out to move from pipeline-ready to production-proven

Slide 17 — Scope: Downstream of PDLC

• Nova governs infra + delivery only; the PDLC (product backlog, code authorship, IDE) is upstream
• Integration is only through the validated contract boundary
• Any upstream source (AI agent, agentic SDLC, dev platform) may produce submissions — all subject to the same compliance standards
• Nova validates the submission, not the author
• Key takeaway: Nova is purpose-built for infrastructure operations, not product development; the scope boundary is clean

Slide 18 — RACI: Who Owns What

• Citizen Developer owns FRs + UAT (via any upstream source — AI agent, SDLC, dev platform — all pass the same gate)
• Platform owns NFRs + infra + QA + prod deploy
• Release Management is co-owned: platform runs attestations agentically, citizen developer oversees + triggers the release (human at stage gate)
• The compliance-standard equivalence is the key: the source does not matter; the submission does
• Key takeaway: you bring FRs + UAT; Nova provides NFRs + infra + QA + prod deploy; the release is co-owned with you at the stage gate

Slide 19 — Production-Grade Guidance via Atelier

• Nova instructs the citizen developer's AI agent via skills (9 markdown files) + an MCP server (4 tools, plugin-registry, stdio)
• The MCP server provides agentic validation beyond deterministic scanners — catches correctness, clarity, observability gaps that Wiz/Checkmarx/Mend cannot
• Atelier is vendored (pinned tag) for audit reproducibility — a validation result is replayable
• This is how Nova ensures the citizen developer's submissions meet production-grade standards regardless of upstream source
• Key takeaway: the citizen developer is not unguided — Nova provides engineering principles via skills + MCP, so every submission meets the same standards

Appendix A1 — Metrics Glossary

• Reference for every metric mentioned in the deck
• Use if the audience asks "what does X mean?"

Appendix A2 — Operating Model & Cost

• The operating cost is negligible (~$0.007/month)
• The zero-cost steady state (D-096 teardown) is the structural mitigation
• References the pre-mortem for the decay-prevention story

---

Slide 20 — 12-Month Product Roadmap

Key takeaway: The next 12 months have a clear product arc — pilot activation → provable trust → compounding ROI → agentic substrate. Each quarter activates one strategic objective.

• This is the product roadmap, not the technical roadmap. The technical milestones (v1.0–v1.19) are behind us; this is forward-looking.
• Q1 (Pilot Activation): re-provision live AWS, activate the first pilot estate, light up the three post-pilot metrics. Onboarding auto-grant ships.
• Q2 (Provable Trust): tamper-evident ledger (Object Lock + JWS), daily checkpoints, live cost reconciliation. Trust is the moat — features can be copied; an immutable decision history cannot.
• Q3 (Compounding ROI + Drift): drift detection + auto-reversal, Infracost→CUR reconciliation on the pilot estate. This is the quarter the CFO points to a number that improves quarter-over-quarter.
• Q4 (Agentic Substrate + Predictive): ML anomaly-forecasting, AI-agent intent surface, multi-cloud preview. The Future Horizon target moves from aspiration to first measurement.
• The roadmap is grounded in the four strategic objectives from the North Star — autonomy, provable trust, ROI, agentic substrate — and the deferred-metric unblock paths from Slide 15.

If asked "what about multi-cloud?": Q4 preview. AWS-only through Q3; Azure/GCP enters preview in Q4. We optimize for depth first, breadth second.

If asked "what about the ML service?": Q4. The predictive-vs-reactive ≥3:1 target requires an ML anomaly-forecasting emitter — the most technically ambitious deliverable on the roadmap.

---

Slide 21 — Quarter-by-Quarter Outcomes

Key takeaway: Each quarter has a concrete deliverable, a target metric, and a strategic-objective grounding. Nothing is hand-waved.

• Q1: three post-pilot metrics go live (Touchless ≥99%, Escalation <0.1%, Accuracy ≥99.5%). The measurement pipeline is already grounded; the denominator activates when the pilot estate runs.
• Q2: Decision Ledger Coverage was already grounded — the tamper-evidence is the Q2 upgrade (SQLite hash-chain → S3 Object Lock + JWS). Cost Savings ≥25% becomes CFO-grade with live CUR reconciliation.
• Q3: Drift Auto-Reversal ≥95% unblocks when the drift scheduler ships. Spend Reduction ≥25% is the same target, now measured against the pilot baseline.
• Q4: Predictive:Reactive ≥3:1 requires the ML forecasting service. AI-Agent Intent Share ≥40% moves from aspiration to first measurement.
• The month-18 destination: "Nova is the layer enterprise leadership points to when they say 'we don't have an infrastructure ops team anymore, and the audit trail is stronger than it ever was.'"

If asked "are these committed or aspirational?": Q1–Q3 are committed (grounded pipeline + known unblock paths). Q4 targets are committed-deliverable, aspirational-metric — the ML service ships, the ≥40% intent share is first measurement (we don't control adoption rate).