DDDeal DossierTransparent acquisition screening
Synthetic sampleSession onlyNo live data

Product case study · Multifamily underwriting

Underwriting you can audit.

Deal Dossier turns a synthetic acquisition case into visible assumptions, deterministic returns, disclosed screening gates, and an evidence-aware diligence posture.

Explore the productSee the system pattern

Illustrative screening, not investment advice. No current property, listing, market, lender, or official-record data is used.

The product

A reviewer sees the logic, not just the answer.

Every result stays attached to the assumptions, timing, provenance state, and fixed gate that produced it.

Private synthetic review candidate
Deal Dossier desktop dashboard showing synthetic acquisition assumptions and deterministic modeled outcomes
Deal Dossier responsive phone dashboard

Scenario proof

One deal. Three explicit cases.

These are recorded outputs from the verified deterministic candidate—not a live calculator and not a forecast.

Illustrative screening posture

Sample clears the illustrative screen

Four of four disclosed gates clear. Verification is still required before any decision.

4 of 4 clear
Year-one cash flow$271,689After modeled operations and debt service
Year-one NOI$932,000Before debt service
Cash-on-cash5.15%Year-one modeled return
Annualized IRR15.86%Across the five-year hold
Equity multiple1.95×Modeled total equity proceeds

How it works

A finance product with a trust contract.

The differentiator is not more data. It is disciplined separation between authored facts, user assumptions, calculations, and missing evidence.

  1. 01

    Shape the sample

    Choose Base, Downside, or Upside, then tune six bounded material assumptions.

  2. 02

    Recompute exactly

    Typed decimal finance logic updates cash flow, debt risk, returns, and sale proceeds.

  3. 03

    Expose every gate

    Four fixed thresholds show precisely why the illustrative screen clears or needs review.

  4. 04

    Verify in diligence

    The output prompts evidence gathering and professional review—never a purchase recommendation.

Transferable engineering

Built for rules-heavy decisions.

Underwriting is the demonstrated domain. The same engineering pattern can support other evidence-sensitive review workflows: structure the inputs, apply explicit rules, explain every finding, and leave judgment with a person.

  1. 01

    Normalize inputs

    Typed, bounded fields turn inconsistent source material into reviewable facts.

  2. 02

    Apply deterministic rules

    The same approved inputs produce the same canonical result every time.

  3. 03

    Attach evidence

    Outputs remain connected to assumptions, provenance, formulas, and thresholds.

  4. 04

    Surface exceptions

    Missing, conflicting, or unverified information remains visible instead of being guessed.

  5. 05

    Keep review human-owned

    The system organizes diligence and escalation; it does not replace accountable judgment.

Boundaries by design

What this project refuses to pretend.

01

No live property claim

The displayed deal, rents, costs, financing, and results are synthetic and illustrative.

02

No hidden persistence

Visitor changes are session-only and reset with the browser lifecycle.

03

No AI-authored returns

Canonical financial outputs come from deterministic typed calculations.

04

No advice or guarantee

Qualified investment, legal, tax, appraisal, engineering, lending, and brokerage review remains required.

August release target

Portfolio-ready. Evidence attached.

Verified candidate293 tests · responsive QA · current digest