One record. Five readers.

Why this matters

Every adjacent product in private markets serves one reader. Sell-side prep tooling serves the operator. CIM platforms serve the banker. Diligence platforms serve the buyer. Covenant monitoring tools serve the lender. LP reporting tools serve the LP. Each tool produces an artifact for its reader; none of those artifacts survive forwarding to the next reader, because each is built on a different framework with different definitions and different evidence standards.

Capital Refinery’s structural claim is that the operator’s self-assessment, the banker’s pitch packet, the buyer’s IC memo, the lender’s credit dossier, and the LP’s verification page are all rendered by the same engine, grading against the same 10-axis framework, with the same deterministic fingerprint. The artifact carries across counterparties because the engine is identical. The only variable is evidence cost — what the reader needed to invest to produce the band they’re looking at.

The five readers

Each reader engages a different artifact rendered by the same engine. The artifact’s evidence level (Self-Assessed → Verification In Progress → Evidence-Confirmed) reflects how the evidence was produced, not which features the engine made available.

The 10 axes, read by all five

The engine grades every business on these ten axes, every time, regardless of which reader is looking. Sub-lane thresholds vary (HVAC vs SaaS vs multi-family vs distribution) but the axes are universal. Two axes — Data Integrity and Financial Consistency — gate the composite verdict for every reader.

Same engine, different evidence levels

The price ladder is calibrated to evidence cost, not feature differences. A $750 Self-Assessment is not the cheap version of the IRA with features stripped. It is the operator-attested version of the same axes — graded by the same engine, against the same thresholds, with the same fingerprint discipline. Document verification costs the operator more. The engine refuses nothing at lower tiers and includes nothing extra at higher tiers; the only variable is the evidence basis behind the band.

• →Self-Assessed ($750) — operator-attested intake; 4-page fingerprinted memo; cannot promote past intake without document ingestion. Forwardable to a CPA, banker, or advisor without platform dependency.
• →Verification In Progress ($4,500 Readiness Gap Review) — document-verified limited scope; the first artifact a sophisticated reviewer treats as evidence rather than attestation.
• →Evidence-Confirmed (Institutional Readiness Assessment, $12,500) — full 10-axis with document evidence behind every axis, plus the verification artifact: a share-token URL a buyer, lender, or LP can independently verify. Distribution is operator-controlled — private by default, shareable with authorized counterparties.
• →Buy-side engagement (institutional — NDA-gated) — same engine running across acquired positions: IC memo, decision lifecycle, covenant forecasting, portfolio readiness rollup, LP verification.

Why the artifact is portable

Three engine primitives make the artifact portable across the five readers:

1. The deterministic fingerprint

Every IRA snapshot, every IC memo, every covenant forecast, every export carries a tamper-evident fingerprint linking the artifact to the engine state at issuance. A reader can confirm independently that the document matches the engine state — no SaaS dependency, no login, no opportunity for post-issuance editing.

2. The share-token verification URL

When the seller, GP, or operator issues a share token, the artifact resolves at a public URL at /p/ira/<token> or /p/<token> for IC memos. The receiving counterparty — buyer, lender, LP, board director — verifies independently. The artifact stands alone after issuance; it doesn’t require ongoing platform access.

3. The refusal to soften

The engine refuses to grade what it cannot observe. Management communication style is opinion, not measurement. Narrative coherence is presentation polish, not evidence. AI-generated sentiment scores on human prose are opinion in numeric clothing. The engine marks unobservable sub-axes as not_observable rather than imputing. That refusal is what makes the artifact credible to sophisticated reviewers — and what makes it portable, because the refusal is identical regardless of which reader is looking.

What this kills

Three operational failures that the same-engine architecture structurally eliminates:

1. Operator-to-buyer handoff drift

A seller produces materials with their banker; the buyer’s IC rebuilds those materials during diligence; the post-close monitoring system rebuilds them again on a different framework. Each rebuild loses fidelity, adds cost, and introduces new opportunities for definitional drift. The same-engine architecture means the seller’s IRA, the buyer’s IC memo, and the post-close monitoring loop all reference the same record. Drift is replaced by inheritance.

2. Banker pitch vs buyer diligence asymmetry

A banker’s CIM tells the story the seller wants told. The buyer’s IC then spends 60–90 days and $150–400K rebuilding the structural verification from scratch. Most of that work is reconciling the seller’s representations against an independent evidence framework — work the IRA already documented. The asymmetry is structural waste. Same-engine architecture compresses the structural portion and frees the buy-side QoE for the deal-specific questions that actually drive decisions.

3. LP review reconstructing from emails

An LP’s ICs reviewing a GP’s fund typically reconstruct the decision chain from emails, board decks, and quarterly letters — none of which were structured to be re-evaluated against an external framework. The same-engine architecture means the LP receives a fingerprinted snapshot per position with sanitized timeline metadata, and the LP’s evaluation framework reads against the same 10 axes that produced the IC memo at decision time. Reconstruction is replaced by inspection.

Where to start, by reader

• •Operators preparing for sale, refinancing, or institutional review — run the $750 Self-Assessment to see the structural blockers a sophisticated reviewer would surface; also institutional readiness for the full IRA methodology.
• •Sell-side bankers running mandates — see for bankers for mandate-cycle integration; refer operators to the SA as a pre-mandate screening tool.
• •Buy-side investment committees underwriting acquisitions — see the diagnostic for the buy-side workspace; Falcon Services Q1 2026 is the canonical buy-side proof case; anatomy of a decision shows one decision end-to-end.
• •Credit officers underwriting facilities — see for lenders for the borrower-side artifact framing; covenant cushion for the buy-side covenant-stress engine.
• •Board directors — see for boards for the fiduciary-infrastructure framing; Decision Timeline + Continuous IC Memo + Anchored Exports cluster.
• •Limited partners reviewing fund governance — see for LPs for the LP-facing framing; LP-sanitized snapshots resolve at /p/<token> for any position the GP has issued a verification URL.

Sources cited

• →ILPA (Institutional Limited Partners Association) Principles — LP governance and reporting standards referenced in the LP-reader framing → https://ilpa.org/principles/
• →SEC Private Fund Adviser Rules (2023 final rule, subsequently vacated by the 5th Circuit in National Association of Private Fund Managers v. SEC, 2024). The rules are not in force; the conceptual framework — quarterly statements, audits, related-party disclosure — continues to inform institutional LP governance expectations independent of the regulatory vehicle. See ILPA Principles above for the active counterpart.
• →NACD (National Association of Corporate Directors) — board governance principles referenced for the Board reader framing → https://www.nacdonline.org
• →Bain Global Private Equity Report — annual industry research on PE deal volume, fundraising, and operational discipline → https://www.bain.com/insights/topics/global-private-equity-report/
• →Cambridge Associates Private Investment Benchmarks — fund performance dispersion data referenced in the LP-reader framing → https://www.cambridgeassociates.com/private-investment-benchmarks/
• →Mauboussin / Lovallo / Sibony on decision quality vs decision outcome — the academic foundation for the same-engine claim's structural argument (referenced in /learn pieces)