Skip to main content
Deal room formatsPDF, XLSX, DOCXYour modelunchanged

Underwrite from your own model

We read the deal room, map every input to the workbook your firm already uses, and hand the numbers back with the source page attached.

Multifamily Acquisition ModelSynthetic sample

Levered IRR

10.6%5-yr

Equity multiple

1.57xhold

Avg cash-on-cash

8.0%avg

Unlevered IRR

7.5%asset

Going-in cap

6.12%Yr 1

Exit cap

6.50%Yr 5

Operating pro forma

Gross potential rentEffective gross incomeNet operating income
$2.54M$1.90M$1.27M$0.63M$0.00M
Yr 1Yr 2Yr 3Yr 4Yr 5
$ / yearYr 1Yr 2Yr 3Yr 4Yr 5
Gross potential rent2,088,0002,150,6402,215,1592,281,6142,350,062
Vacancy and credit loss(125,280)(129,038)(132,910)(136,897)(141,004)
Effective gross income1,962,7202,021,6022,082,2502,144,7172,209,059
Operating expenses(824,000)(844,600)(865,715)(887,358)(909,542)
Net operating income1,138,7201,177,0021,216,5351,257,3591,299,517
Debt service(695,175)(695,175)(695,175)(695,175)(695,175)
Levered cash flow443,545481,827521,360562,184604,342

Assumptions

from your model

Property and operating

Units120
Avg monthly rent / unit1,450
Vacancy and credit loss6.0%
OpEx / unit / yr6,867
Rent growth3.0%
Expense growth2.5%

Financing and exit

Purchase price18,600,000
Loan-to-value65%
Interest rate5.75%
AmortizationInterest only
Hold period (years)5
Exit cap rate6.50%
Selling costs1.5%

Your assumptions, your line items, and the outputs your committee already reads.

Synthetic deal and illustrative assumptions. Your model, your numbers and your logic stay yours; the system fills them in.

One process, three chapters

From the first screen to the committee, without re-keying

Screening, modelling and the memo were three tools and three re-keyings of the same numbers. Here they are one process, and every figure keeps a line back to the document it came from.

IInitial screening

Nothing unqualified reaches an analyst.

Every inbound deal, OM and broker email is read, extracted and run against your hard disqualifiers before a human opens it. What survives arrives as a structured snapshot: the numbers, where each one came from, and why it passed.

IIPro forma

An auditable model, not a black box.

Operating statement, sources and uses, debt schedule, returns and sensitivity, in the workbook your firm already uses. Comparable overlays carry their source. You can trace any cell back to the document it came from.

IIIInvestment committee

The recommendation is drafted. The decision stays yours.

Assumptions, scenarios, risks and a price ceiling assembled into a memo in your committee format, with the evidence attached to each claim. The system never issues a verdict; it makes the case reviewable in an afternoon instead of a week.

Deal room · to your model
01

Drop it in

  • Rent roll.xlsx412 KB
  • T12 statement.pdf1.1 MB
  • Offering memo.pdf8.4 MB
  • Leases (14).zip22 MB
02

Map to your workbook

  • Avg in-place rent

    Assumptions!B7

  • Unit count

    Assumptions!B4

  • Vacancy

    Assumptions!B9

  • OpEx by line

    OpEx!C4:C19

3 of 4 mapped

03

Read the output

  • First screenPass
  • Pro forma5 yr
  • IC memoDraft

Every figure links to the page it was read from.

The three steps of the real intake, in the order the work happens.

Synthetic deal room. The step set is our live process.

Why it is auditable

Every extracted field keeps the page it was read from

An extraction you cannot check is a number you have to re-key to trust, which is the work the system was supposed to remove. Each field carries the document and the page it came from, so a reviewer verifies in a click instead of reopening the deal room.

Evidence · rent roll

Source document · page 3

RENT ROLL, as of 30 Jun

UnitTypeRentTerm
A-1012BR / 1BA1,39512 mo
A-1022BR / 1BA1,45012 mo
A-1031BR / 1BA1,150MTM
B-2013BR / 2BA1,79524 mo
B-2022BR / 1BA1,42512 mo

Extracted fields

  • Unit count120

    read from p.1, header

  • Avg in-place rent$1,450

    read from p.3, col. D

  • Lease term, B-20124 months

    read from p.3, row 4

  • Vacancy at scan6.0%

    read from p.1, summary

Each field, its value, and the page reference it was read from.

Synthetic extraction. Page references are in the real format.

Questions about underwriting

Do we have to change our model?

No, and that is the point of the design. Your model encodes how your firm decides, and replacing it would replace your judgement with ours. The system reads the deal room and puts the line items where your existing workbook expects them.

Does the panel recompute when we change an assumption?

The figures on this page do not recompute; they are a picture of a model surface, and the caption says so. In the delivered system the workbook is your own, so recomputation is whatever your workbook already does when a cell changes. We are deliberately explicit about this because a marketing panel labelled live that is not live is the easiest dishonest claim in this category to make.

What can it read?

Offering memoranda, rent rolls, operating statements, loan documents and broker emails, as PDF, XLSX or DOCX. Where a document is a scan, it is processed as one and flagged as lower confidence rather than passed through as if it were text.

Does it make the investment decision?

No. It drafts the case and attaches the evidence to every claim in it. The committee decides. A system that issued verdicts would be asking your firm to delegate the one thing it exists to do.

How do we start?

With a proof of concept: $3,000, two weeks, run on one of your own live deals in your own model. You see the screen, the pro forma and the memo it produces from a deal you already know the answer to, which is the only honest way to judge it.

See it run on one of your own deals.

Bring one live deal and your model. We show you the screen, the pro forma and the memo it produces.