Neutral by construction. Contributor-anonymous.

Private market intelligence compounds across firms, not within them.

A mid-market allocator makes a few dozen decisions a year and waits seven to twelve years to learn whether each one was right. No single firm ever accumulates enough labelled history to turn AI into an investment edge.

TrancheBook operates the layer no single firm can build — and the tooling that pays for itself at a network size of one.

For accredited investors and institutional allocators only. TrancheBook is a financial publisher, not a broker-dealer.

Firm intelligence today, network intelligence next

Firm intelligence

Available now

A workflow product worth buying with no network attached. Your own private Decision Memory captures thesis, assumptions, risk flags, internal dissent, approval conditions and decline rationale — at the moment of decision, before the outcome exists to contaminate it.

IC-ready research

Standardized packs that separate facts, assumptions, judgments and unknowns — every claim linked to its source.

Adversarial review

A red-team pass builds the strongest counter-case to every recommendation before it reaches committee.

Decision Memory

Post-decision monitoring wired back to the original thesis, so drift is visible rather than remembered.

The pooled Decision Memory

In development — founding cohort forming

Structured contribution is produced as a by-product of the Layer 1 workflow — capture is not an additional task. As the founding cohort forms, members will receive pooled intelligence in return for contribution: how situations of a given shape resolved across the network, which risk flags actually preceded losses, what the network declined and what happened to those opportunities afterwards.
  • • Contributor-anonymous and subject-named — the network never sees which member declined
  • • Give-to-get, scored and audited, not honour-system
  • • Derived schema fields only — never third-party documents, LP identities, or anything received under NDA

The trust infrastructure

The advantage isn't that we use AI — every competitor will. It's the verification stack around the AI: deterministic kill gates, a public calibration record, and audit-grade source provenance on every claim.

We reject deals that don't fit

Every memo passes deterministic kill gates before it ships. An LBO request on a $300B GSIB will be turned down — and the reasoning logged.

Memos carry a public accuracy track record

Every prediction is resolved against eventual reality and published as a calibration retrospective on the blog. We are willing to be wrong in public — that's what makes the track record real.

Every claim links to its source

EDGAR filings, S-1s, named news sources, GDELT entries — every assertion in a memo is hyperlinked to its origin. No black-box reasoning, no synthesis without a citation.

Revenue stacks in the order the layers mature

Firm intelligence is sold today. Network membership is priced as a premium tier over it and gated on contribution — terms are being set with the founding cohort.

Layer 1 — Firm intelligence
Enterprise subscription, anchored to our current institutional research pricing of $25,000–$75,000 ACV. Firm-intelligence tier pricing is being finalized with anchor members.
  • • Private Decision Memory
  • • IC-ready research packs with source provenance
  • • Adversarial red-team pass on every recommendation
Layer 2 — Network membership
A premium tier over Layer 1, gated on contribution volume and quality. Pricing is directional until anchor terms are agreed with the founding cohort.
  • • Access conditioned on audited contribution
  • • Contributor-anonymous, subject-named outputs
  • • Opens once the founding cohort is in place

Stay informed

Calibration retrospectives, research notes, and updates as the founding cohort forms — straight to your inbox.

Financial publisher status

TrancheBook is a financial publisher and research platform — not a broker-dealer. Indications of interest submitted on the platform are routed to our registered broker-dealer partner, Rainmaker Securities, on a flat-fee basis. Our research and scoring outputs are architecturally walled off from broker fee economics. Please review our risk disclosures before proceeding.