§ 00 ── POST-PMF COMPANY OS · SERIES D+
Preserve methodology
across leadership transitions.
For Series D+ companies. 200+ employees. $30M+ ARR. Institutional governance is first-class. Board members, investors, regulators demand audit-grade decision provenance. Multiple leadership transitions over the company's life. The substrate preserves what would otherwise dilute. Per Diana Hu's YC framing: incumbents that successfully go AI-native typically spin up small internal teams that build AI-native systems from scratch separate from the core business (the Mutiny pattern). The substrate is the platform that team builds on — and the 12-agent canonical fabric (expanded to 14–17 at D+ via sector-specific Class A additions per Customization Decision Tree) is the operational layer of that platform.
§ 01 ── INSTITUTIONAL REALITY
The substrate at this stage is institutional infrastructure.
By Series D+ you've experienced multiple VP transitions. You've fielded investor inquiries that required showing decision provenance. You've prepared (or are preparing) for SOC 2 Type II, sector-specific compliance, or M&A diligence. The audit chain you have is fragmented across tools. Methodology lives partially in current leadership heads, partially in shared docs of varying age, partially nowhere. Every change to a core process risks breaking something that already works — which is why the skunk-works pattern (small internal team building the AI-native version in parallel with the running business) is the canonical entry path at this stage. Phase 0 (3 weeks, free) establishes the operating envelope for that team. The canonical 9-seat Methodology Council (expanded from 7-seat at B/C standard) ratifies the substrate's operational integration with existing governance.
Removing the substrate at Series D+ wouldn't be like removing a tool. It would be like removing a load-bearing piece of how the company operates.
§ 02 ── METHODOLOGY PRESERVATION ACROSS LEADERSHIP
Multiple VPs cycle through. Methodology has to survive each transition.
Statistically, a Series D+ company will see 2–4 VP transitions over the next five years. Each transition risks loss of methodology that lived in the departing leader's head. The substrate's job: preserve methodology — encoded in L2 overlay (customer-owned · MSA § 8.3 · architecturally enforced via F-FW1 at operator-tier) — so each transition doesn't reset institutional memory.
01 · CURRENT LEADERSHIP METHODOLOGY
Captured continuously into the L2 overlay.
Methodology of every current functional leader captured continuously into the L2 overlay via Knowledge Curator (Class A · T2 · canonical · extended methodology-capture mode at D+ scale). Each VP's decision heuristics, operating principles, and communication patterns are made explicit — not left as oral tradition. Per-VP voice corpus calibrated (κ ≥ 0.85 Cohen's weighted · verified by Eval Suite Runner · Class B · T4).
02 · DEPARTING LEADERSHIP PRESERVATION
Structured capture sprint at announcement.
When a leader announces departure a structured capture sprint runs (operator-facilitated · 1–2 weeks). It captures their accumulated methodology before they leave via Knowledge Curator + Decision Brief Drafter + Strategy Lead joint operation. Their L2 overlay section becomes the institutional memory that outlasts their tenure. Audit-emitted via Audit Curator (F4 sole-emit · L0 doctrine · NEVER-deletable on critical).
03 · INCOMING LEADERSHIP ONBOARDING
Replay the predecessor's reasoning with full audit chain.
A new VP can query: "how did my predecessor think about X?" The substrate replays the predecessor's reasoning with full audit chain context (D7 §12 Step 02 REPLAY · ≤15-minute SLA). New leaders build on institutional memory rather than restarting from scratch. Replay-test on every model swap verifies structural continuity across model generations.
FIG.01 — Five-year horizon · VP transitions · L2 overlay continuous across departures · canonical capture via Knowledge Curator (Class A) + audit chain via Audit Curator (Class B · F4 L0 doctrine).
FIG.01 — Five-year horizon · VP transitions · L2 overlay continuous across departures · canonical capture via Knowledge Curator (Class A) + audit chain via Audit Curator (Class B · F4 L0 doctrine).
§ 03 ── STANDING VERIFICATION RIGHTS
Stakeholders exercise D7 §12 verification on demand. Canonical 5-step protocol.
Board members, Series D+ class investors, regulators (where applicable to your domain), and acquirers (during M&A diligence) all hold standing inspection authorization under the D7 §12 canonical 5-step protocol per L0 doctrine. Quarterly cadence with investors and board is typical. Annual comprehensive audits. M&A diligence inspections. Regulatory verifications.
D7 §12 · STANDING VERIFICATION RIGHT · CANONICAL 5 STEPS · BOARD · INVESTORS · REGULATORS · ACQUIRERS · METHODOLOGY COUNCIL SEATS
STEP
NAME
NOTICE
WHAT IT VERIFIES
01
INSPECT
IMMEDIATE (CUSTOMER) · 3–5 BD (OPERATOR · ANONYMIZED)
Read access to substrate + audit chain. Any agent contract (14–17 at D+), any audit chain entry, methodology overlay, L8 floor implementations, 9-seat Methodology Council decisions log, KPI evidence bundles.
02
REPLAY
IMMEDIATE · ≤15 MIN SLA
Re-execute any past decision via audit chain query. Audit Curator + Knowledge Curator provide queryable interface. Multi-year audit chain depth (2–4 years at D+) means decision archaeology compounds.
03
VERIFY
IMMEDIATE
Confirm code-path absence for each L8 floor. L8-Enforcer's quarterly self-test logs reviewable. All applicable floors PASS expected (typically 8–9 floors active at D+ · F1–F6 canonical + F7–F9 optional per pattern). Pressure-test refusal traces available.
04
CROSS-CHECK
IMMEDIATE
Verify κ against 9-seat Methodology Council-ratified baseline. Eval Suite Runner re-computes κ on sampled decisions. Tolerance ±0.02 PASS · ±0.05 WARNING · beyond CRITICAL.
05
ESCALATE
72H CRITICAL · 1 WK WARNING
Route critical findings to operator-tier recursive anchor (TQC Methodology Council · 2-of-2 founding seats). Authority on bypass-discipline refusal, floor breach, F4 sole-emit violation, κ collapse.
FIG.02 — D7 §12 canonical 5-step protocol per L0 doctrine · standing authorization for board, investors, regulators, acquirers, Methodology Council seats · pass criteria binary · architecturally cooperative.
FIG.02 — D7 §12 canonical 5-step protocol per L0 doctrine · standing authorization for board, investors, regulators, acquirers, Methodology Council seats · pass criteria binary · architecturally cooperative.
§ ── METHODOLOGY DECAY · ESTIMATE
What's leaking right now.
Three inputs. A doctrinal estimate of un-captured substantive decisions per quarter. Conservative on purpose.
UN-CAPTURED · ESTIMATE
28,080
SUBSTANTIVE DECISIONS · HORIZON WINDOW
RISK SCORE
100/100
DOCTRINAL · CONSERVATIVE
TIER
HIGH
PHASE 0 RECOMMENDED · 7 DAYS
§ 04 ── L8 FLOORS · AUDIT CHAIN · COMPLIANCE
Audit chain depth multi-year. L8 floors absolute. Compliance posture mature.
AUDIT CHAIN · MULTI-YEAR DEPTH
By Series D+ engagement, audit chain history is 2–4 years deep. Decision archaeology (via Decision Brief Drafter audit-emissions captured by Audit Curator F4 sole-emit) becomes load-bearing for D7 §12 verifications. Replay against any point in the past is structurally deterministic (LLM-stochastic at narrative level · structurally identical per replay-test infrastructure · per L0 §4.2 replay-test outcome doctrine).
FIG.03 — Audit chain depth compounds quarter over quarter · canonical quarterly D7 §12 cadence at D+ scale (Q1 · Q2 · Q3 · Q4) · structurally deterministic replay.
FIG.03 — Audit chain depth compounds quarter over quarter · canonical quarterly D7 §12 cadence at D+ scale (Q1 · Q2 · Q3 · Q4) · structurally deterministic replay.
L8 FLOORS · CANONICAL F1–F9 · ABSOLUTE AT D+
All applicable canonical L8 floors enforced architecturally at D+. Code paths for autonomous commits in F1–F9 do not exist. L8-Enforcer (Class B · T4 · recursively self-protected · cannot be silenced) runs quarterly self-tests (typically 8–9/9 floors PASS at D+ scale).
FL
FLOOR
TYPICALLY ACTIVE AT D+
F1
Money-Movement
YES — any revenue or spend handling
F2
Customer Data
YES — always · L0 doctrine adjacent
F3
Compliance / Risk
YES — multi-jurisdictional regulatory
F4
Audit Sole-Emit · L0 DOCTRINE
YES — always applies
F5
Cohort (k ≥ 5)
YES — cross-team aggregation
F6
Code-Shipping
YES — D+ typically regulated
F7
Unregulated alt
typically NO at D+ (F6 preferred)
F8
Multi-Tenant Isolation
YES if multi-tenant SaaS
F9
Employee Privacy
YES — GDPR + global workforce
All applicable floors PASS quarterly (typically 8–9/9 at D+ per Customization Decision Tree). Sector-specific calibrations (regulated fintech, healthcare, pharma) may extend floor catalog per L1 framework.
COMPLIANCE POSTURE · MATURE AT D+
SOC 2 Type II audit-aware architecture (audit chain provides CC6/CC7/CC8 evidence by default). EU AI Act Article 12 record-keeping by default. NIST AI RMF operationalized via 4-phase engagement + 9-seat Methodology Council monthly cadence. GDPR right-to-erasure cascades (24h cascade · 72h purge per F2 + F9). HIPAA-eligible deployment for healthcare (with BAA · default L8 catalog amended for PHI). Sector-specific frameworks where applicable (regulated fintech: PSA · APPI · JFSA · PCI DSS per L1 Compliance Mapping Library). ISO 27001 A.5 · A.8 · A.12 · A.18 controls map to F1–F9 floor catalog.
§ 05 ── THE ENGAGEMENT · EXTENDED
Extended timing. Substantial footprint. Standing engagement.
Phase 0 extends to 3 weeks because discovery includes founders, full functional leadership, governance representatives, and multi-jurisdictional regulatory mapping. Phase 1 extends to 12–14 weeks because the canonical 12-agent base + 0–5 sector-specific Class A additions activate progressively. KPI gate at canonical Week 9 leaves 3–5 weeks for D7 §12 verification readiness + optional remediation. Phase 2 carries substantial retainer footprint matching governance-grade operation. Phase 3 quarterly residency is the standing default.
FIG.04 — Engagement ribbon · extended timing for governance-grade operation · KPI gate at canonical W9 · Phase 3 quarterly residency standing default.
FIG.04 — Engagement ribbon · extended timing for governance-grade operation · KPI gate at canonical W9 · Phase 3 quarterly residency standing default.
PHASE 0
Extended discovery with founders + full functional leadership + governance representatives (Days 1–5). Brand DNA Step 0 research (Days 4–10 · canonical DEEP depth · 40–50+ sources · multi-jurisdictional regulatory mapping included). Methodology Council Workshop (Days 10–14 · canonical 9-seat Council expansion at D+ · adds Multi-Jurisdictional Representative + Audit Liaison Representative to standard 7-seat · charter ratified · κ baseline lock · quorum 6-of-9 monthly · 7-of-9 MAJOR · 8-of-9 floor amendments). Comprehensive methodology v0.1 capture. D7 §12 inspection drill (Days 15–18). Synthetic-data PoC + GO/NO-GO at end of Week 3. If NO-GO, customer keeps all artifacts unconditionally per MSA § 8.3.
PHASE 1
Canonical 12-agent base + 0–5 sector-specific Class A additions integrated. Sector additions per Customization Decision Tree: People Steward (F9 + F3), External Comms Steward (F2 + F3), Audit Liaison (F4 + F3), and sector-specific Drafters (e.g. Clinical Operations Lead for healthcare, Trading Operations Lead for fintech). Total: 9–12 Class A + 5 Class B = 14–17 agents at D+. Class B canonical 5 stable across all stages (Audit Curator · L8-Enforcer · Registry Maintainer · Eval Suite Runner · Version Controller). D7 §12 verification readiness exercised (canonical 5 steps rehearsed). Quarterly L8 self-test rehearsed in advance. 8 universal KPIs (K1–K8) + 1–2 customer-specific (K9/K10) gate at canonical Week 9. 6+ of 8 universal PASS → Phase 2 GO at end of Phase 1 (W12–W14 after optional remediation).
PHASE 2
Substantial retainer footprint matching governance-grade operation. 9-seat Methodology Council monthly + ad-hoc + quarterly + annual cadence. Quarterly D7 §12 exercises included (canonical 5 steps operationally validated each quarter). Annual κ re-baselining (8-of-9 Council quorum). Multi-year engagement typical (5+ years common at D+ scale). Replay-test infrastructure operational on every model swap (structural fingerprint discipline · per L0 §4.2 outcome doctrine).
PHASE 3
Standing default at D+. Evolution work each quarter. Loop accretion toward Year-2 ~25 agents across 6 loops (canonical 6-loop sequence per L1 framework). Pattern migration L2 → L1 candidates surfaced (concurrence-gated per F-FW4 + MSA § 8.4 + customer-attribution + compensation). Stage Transition Runbook (D+ → IPO/M&A) activates for IPO readiness or M&A diligence — substrate becomes governance evidence for prospectus + decision archaeology compounds for D&O coverage. Cohort intelligence activates if cohort reaches N≥5 (D+ customers often the anchor for cohort activation given multi-year engagement depth).
§ 06 ── INSTITUTIONAL INFRASTRUCTURE
The substrate is what the company runs on.
By Series D+ the substrate is preserving methodology across leadership transitions that would otherwise destroy institutional knowledge (per § 02). The substrate is supporting D7 §12 verifications that would otherwise be impossible to execute (per § 03 · canonical 5 steps). The substrate is carrying audit chain history that would otherwise live in fragmented systems (2–4 years deep per § 04). The substrate is supporting governance-grade operation that would otherwise require dedicated teams of governance-and-audit-ops specialists. The skunk-works team that started inside the incumbent gradually becomes the integrator the org chart used to be. The 14–17 agents (canonical 12 + sector additions) operate under one customer-tier audit chain · F4 sole-emit · κ ≥ 0.85 · D7 §12 inspection-ready.
For most Series D+ customers, the substrate is not "AI assistance." It is institutional infrastructure — per Diana Hu's YC framing of AI-native operating companies + Jack Dorsey's humans-at-the-edge framing at Block, operationally realized at governance grade.
§ 07 ── UNDER THE HOOD · NDA-TIER
D+ -specific depth available pre-engagement.
What's public on this page: canonical 12-agent base + sector-specific Class A additions at D+ · 12–14 wk Phase 1 with W9 KPI gate · 9-seat Methodology Council · canonical D7 §12 5 steps · F1–F9 typically active · standing Phase 3.
What's available under NDA (recommended at D+ given M&A, IR, regulator-adjacent sensitivity):
- 01
Customization Decision Tree applied to your D+ pattern — which sector-specific Class A additions activate, which optional floors (F7 if unregulated subset · F8 multi-tenant · F9 always at D+) active.
- 02
Pricing range for Phase 1 fixed-fee + Phase 2 monthly retainer at D+ scale (enterprise band · governance-grade · negotiated per engagement · specifics discussed at Phase 0).
- 03
9-seat Methodology Council seat composition recommendations + workshop facilitation playbook.
- 04
Brand DNA research scope at DEEP depth (40–50+ sources · multi-jurisdictional regulatory mapping included).
- 05
D7 §12 drill format · pressure-test procedures · cross-tier composition verification · escalation routing · board / investor / regulator / acquirer-facing protocols.
- 06
Compliance Mapping Library applied per jurisdiction (GDPR · HIPAA · SOC 2 · ISO 27001 · NIST AI RMF · EU AI Act · PSA · APPI · JFSA · FCA · sector overlays).
- 07
Stage Transition Runbook D+ → IPO/M&A mechanics · substrate-as-governance-evidence for prospectus · D&O coverage adequacy · audit-ready financial controls.
- 08
Pattern migration L2 → L1 mechanics · concurrence-gated · attribution + compensation per MSA § 8.4 + F-FW4 architectural floor at operator-tier.
- 09
Worked examples at D+ scale under NDA (additional anonymized customer studies in regulated fintech, pharma, healthcare, multi-jurisdictional SaaS).
- 10
Cohort intelligence activation at N≥5 (cross-engagement κ benchmarks · regulatory pattern propagation · decision archaeology aggregation).
§ ── NOT YOUR STAGE
The substrate is one architecture. The conversation calibrates per stage.
§ END ── GOVERNANCE-GRADE
If you operate a Series D+ company — Phase 0 is 3 weeks, free. Governance-grade by design. You keep all artifacts either way (per MSA § 8.3 · architecturally enforced via F-FW1 at operator-tier).
At a different stage?