New deal
Upload the offering memorandum. The model tests the downside and writes the workbook and the memo, in place of a blank spreadsheet.
Drop the offering memorandum here
PDF or .pptx up to 4 MB and 600 pages, Excel or CSV up to 32 MB.
PDF, PowerPoint, Excel, CSV, or text.
Deal desk
Ithaca, NY, Mixed-Use
Recorded public records for the Ithaca example. The recorded view does not write anything into this deal.
Recommendation to investment committee
Project IRR 15.34% clears the 15.00% hurdle. Min DSCR 1.46x stays above the 1.00x floor.
Decided on Project IRR 15.34%. Clears the 15.00% hurdle by 0.34 pts.
Source: engine pro forma, NOI by year.
| Year | NOI | Note |
|---|---|---|
| Y1 | $594,066.40 | |
| Y2 | $612,661.89 | |
| Y3 | $631,834.59 | |
| Y4 | $651,602.28 | |
| Y5 | $671,983.33 | |
| Y6 | $692,996.62 | |
| Y7 | $714,661.67 |
CAGR 3.13%
Source: exit cap grid at this deal's rent growth.
| Exit cap | Levered IRR | Note |
|---|---|---|
| 5.75% | 18.27% | |
| 6.00% | 17.26% | |
| 6.25% | 16.29% | |
| 6.50% | 15.34% | Base case |
| 6.75% | 14.42% | |
| 7.00% | 13.51% | |
| 7.25% | 12.63% |
15.00% hurdle
Breakeven 6.59%
Source: equity cash flows, cumulative cash returned.
| Year | Multiple | Note |
|---|---|---|
| Y0 | 0.00x | |
| Y1 | 0.06x | |
| Y2 | 0.12x | |
| Y3 | 0.19x | |
| Y4 | 0.27x | |
| Y5 | 0.35x | |
| Y6 | 0.44x | |
| Y7 | 2.4351x | Payback year |
1.0x payback
Source: senior coverage by year.
| Year | DSCR | Note |
|---|---|---|
| Y1 | 1.46x | Weakest year |
| Y2 | 1.50x | |
| Y3 | 1.55x | |
| Y4 | 1.60x | |
| Y5 | 1.65x | |
| Y6 | 1.70x | |
| Y7 | 1.75x |
1.00x floor. 1.20x lender floor
Clears the 15.00% hurdle by 0.34 pts.
IRR meets the 15.00% hurdle at a 6.59% exit cap, 9 bps wider than today.
50 bps wider: IRR 13.51%. 50 bps tighter: IRR 17.26%.
IRR meets the 15.00% hurdle at 2.87% rent growth, 0.13 pts slower than today.
100 bps slower: IRR 12.68%. 100 bps faster: IRR 17.81%.
Years 1 to 7. Excludes $6.1M net sale proceeds in Year 7.
NOI over annual debt service. Floor 1.00x.
Green clears the 15.00% hurdle, red does not. Base case outlined.
| Exit cap by rent growth | 1.50% | 2.00% | 2.50% | 3.00% | 3.50% | 4.00% | 4.50% |
|---|---|---|---|---|---|---|---|
| 5.75% | 14.43% | 15.75% | 17.03% | 18.27% | 19.46% | 20.63% | 21.76% |
| 6.00% | 13.35% | 14.70% | 16.01% | 17.26% | 18.48% | 19.66% | 20.80% |
| 6.25% | 12.30% | 13.68% | 15.01% | 16.29% | 17.52% | 18.72% | 19.88% |
| 6.50% | 11.26% | 12.68% | 14.04% | 15.34% | 16.60% | 17.81% | 18.99% |
| 6.75% | 10.25% | 11.70% | 13.09% | 14.42% | 15.69% | 16.93% | 18.12% |
| 7.00% | 9.25% | 10.74% | 12.16% | 13.51% | 14.81% | 16.07% | 17.28% |
| 7.25% | 8.27% | 9.79% | 11.24% | 12.63% | 13.95% | 15.23% | 16.46% |
1 more model limitation applies: Returns are pre-tax and pre-fee. Read the model limitations.
Buy box
Edit buy boxFits
A second opinion on the broker package. Each figure the offering memo states is checked against the rent roll and operating statement you upload, and against free public records saved on this deal. Every status cites its source.
No offering memo figures are on this deal yet. Read an offering memo on the desk, or enter the memo figures in the OM Truth Score panel below.
The file is read in your browser. Only the totals are kept on the deal: units, occupied units, average rent, income, expenses and taxes.
Broker claims against the T-12, the rent roll, and public records on this deal. A missing source is Not checked. It is not treated as a discrepancy.
Not scored
No claim was checked, so there is no score. Missing evidence does not lower it.
The OM Truth Score is 100 times the weighted credit of the claims that were actually checked, divided by the weight of those claims. Weights: NOI 30, pro forma rent 15, occupancy 10, expense ratio 10, taxes 10, units 10, square footage 5, year built 5, cap rate 5. A claim with no evidence is Not checked. Its weight is left out. Missing evidence is not a discrepancy and it does not lower the score. If nothing was checked, the score is not scored. Credit for a checked claim is 1 minus the normalized gap, and it cannot go below 0. The family's credit is the lowest credit among the evidence rows for that claim. Money, rent, taxes, expense ratio, and square footage agree within 2 percent. Past that, credit falls to 0 across the next 20 percent of relative gap. Occupancy agrees within 1 percentage point. Past that, credit falls to 0 across the next 15 percentage points. Units agree only when the counts are equal. Credit is 1 minus the unit gap divided by the larger count. Year built agrees within one year. A wider gap takes the whole credit. It is not converted into NOI. Cap rate agrees within 15 basis points. Past that, credit falls to 0 across the next 150 basis points. HUD Fair Market Rent and Census ACS median gross rent are benchmarks. They are shown and they are not scored. They are not comps. Open violations are an exposure. They are shown and they are not scored, and they are not converted into NOI. A positive LL97 penalty estimate is an exposure. It is not scored. It is subtracted in the adjusted NOI because a historical T-12 does not contain a future penalty. Adjusted NOI starts from the T-12 NOI when one is present, otherwise from the OM NOI. A tax assessment above the OM tax is subtracted only when the starting NOI is the OM figure, because a T-12 already includes the taxes that were paid. The adjusted case runs on a copy. The base verdict does not change unless that copy is applied. Dollar impact on NOI is the claim minus the evidence for income, and the evidence minus the claim for taxes. Value impact divides that NOI gap by the OM cap rate, or by the model's entry cap when the OM states no cap rate. The cap-rate row repeats the NOI gap capitalized at the claimed cap. It is the same dollars, not a second deduction.
T-12 NOI not uploaded. Rent-roll NOI not computed. The rent-roll figure is in-place residential rent times occupied units, minus T-12 operating expenses, and only when all three are present. It is not substituted for the adjusted NOI.
Claim Not stated. Evidence Not stated.
Uploaded T-12 · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state a pro forma NOI. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
Uploaded rent roll · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state a pro forma rent. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
HUD Fair Market Rents · Not on the snapshot · Not checked
NOI impact not converted. Value impact not converted.
Not checked. No HUD one-bedroom Fair Market Rent is on the intel snapshot. A benchmark is not a comp, and it is not scored.
Claim Not stated. Evidence Not stated.
Census ACS · Not on the snapshot · Not checked
NOI impact not converted. Value impact not converted.
Not checked. No Census median gross rent is on the intel snapshot. It is not scored.
Claim Not stated. Evidence Not stated.
Uploaded T-12 · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state an occupancy. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
Uploaded T-12 · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state an expense ratio. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
NYC assessment or T-12 · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state taxes. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
Uploaded rent roll · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state this figure. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
Uploaded rent roll · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state this figure. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
PLUTO · Not on the snapshot · Not checked
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state a year built. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
Uploaded T-12 · Uploaded document · user-upload · Date not stated
NOI impact not converted. Value impact not converted.
Not checked. The OM did not state a cap rate. A missing claim is not a discrepancy.
Claim Not stated. Evidence Not stated.
NYC violations · Not on the snapshot · Not checked
NOI impact not converted. Value impact not converted.
Not checked. No HPD, DOB, or ECB violation count is on the intel snapshot. A missing count is not zero, and it is not scored.
Claim Not stated. Evidence Not stated.
LL84 and LL97 · Not on the snapshot · Not checked
NOI impact not converted. Value impact not converted.
Not checked. No LL97 penalty estimate is on the intel snapshot. It is not scored, and a missing estimate is not zero.
No adjusted NOI yet. It needs an OM NOI or a T-12 NOI.
New York. Millage on price.
Not flagged
| Metric | Before | After |
|---|---|---|
| Real estate tax | $95,000$95,000.00 | $95,000$95,000.00 |
| Year 1 NOI | $594,066$594,066.40 | $594,066$594,066.40 |
| Minimum DSCR | 1.46x | 1.46x |
| Levered IRR | 15.34% | 15.34% |
| Recommendation | PURSUE | PURSUE |
Source: Outside New York City, New York does not reassess to the sale price in this check. Post-sale tax here is your millage times the purchase price. This is an estimate, not a tax bill.
No sale-reset rule is applied. The millage starts at the seller's implied rate, 1.1176%. Post-sale tax is that rate times the purchase price until you edit it. Later years in the scenario grow with the deal's expense growth. The engine keeps one expense growth rate.
The base recommendation on this desk does not change. The after column is a scenario on a copy.
Editable assumption. 2.00 means the tax estimate is 2.00% of the purchase price. Clear the field to use the jurisdiction rate.
The highest price at which a loan sized to the LTV cap, the DSCR floor, and the debt-yield floor still meets every hurdle. Solved by bisection over the same engine as the desk. The price is not written onto the deal unless you apply it.
$8,862,491
The price is set by Levered IRR. The loan is set by DSCR.
Metrics follow the property type. Same engine inputs underneath.
What builds the IRR
Without the 49 bps of cap compression (6.99% entry to 6.50% exit), IRR is 13.55%, 1.45 pts under the 15.00% hurdle.
| Step | How it is run | Change | Running total |
|---|---|---|---|
| No thesis | Growth at zero, sold at the 6.99% entry cap | total | 6.22% |
| NOI growth | Sale at the entry cap, growth as modeled | 7.33 pts | 13.55% |
| Exit cap | 6.99% entry to 6.50% exit | 2.13 pts | 15.68% |
| Interaction | Growth priced at the lower cap | (0.34 pts) | 15.34% |
| As underwritten | The desk's own run | total | 15.34% |
Source: Underwrite model, illustrative case
The same outputs, read the way a senior lender sizes a loan.
No input on this deal was read from an offering memo in this session. Every figure below is the engine's own output for the deal on screen.
Scheduled rent and other income, less vacancy on each line.
Engine: buildProForma, lib/engine/proforma.ts
Inputs: Apartment units, Average monthly rent, Residential vacancy, Commercial area, SF, Commercial rent per SF, Commercial vacancy, Other income.
Year 1 effective gross income less operating expenses. Management and reserves are charged as a share of effective gross income.
Engine: buildProForma, lib/engine/proforma.ts
Inputs: Apartment units, Average monthly rent, Residential vacancy, Commercial area, SF, Commercial rent per SF, Commercial vacancy, Other income, Real estate taxes, Insurance, Management, share of EGI, Repairs and maintenance, Utilities, Marketing and G&A, Reserves, share of EGI.
Year 1 NOI divided by the purchase price.
Engine: buildReturnsSummary, lib/engine/returns.ts
Inputs: Purchase price, Apartment units, Average monthly rent, Residential vacancy, Commercial area, SF, Commercial rent per SF, Commercial vacancy, Other income, Real estate taxes, Insurance, Management, share of EGI, Repairs and maintenance, Utilities, Marketing and G&A, Reserves, share of EGI.
Purchase price times loan to value.
Engine: buildDebtSchedule, lib/engine/debt.ts
Inputs: Purchase price, Loan to value.
Total uses (price, closing costs and loan fees) less the loan.
Engine: buildSourcesUses, lib/engine/sourcesUses.ts
Inputs: Purchase price, Closing costs, Loan to value, Origination fee.
Each year's NOI divided by that year's loan payments; the lowest year of the hold.
Engine: buildReturnsSummary, lib/engine/returns.ts, on buildDebtSchedule, lib/engine/debt.ts
Inputs: Purchase price, Loan to value, Interest rate, Amortization, Hold, Apartment units, Average monthly rent, Residential vacancy, Commercial area, SF, Commercial rent per SF, Commercial vacancy, Other income, Real estate taxes, Insurance, Management, share of EGI, Repairs and maintenance, Utilities, Marketing and G&A, Reserves, share of EGI, Rent growth, Expense growth, Other income growth.
The rate that sets the equity cash flows to zero: equity in at closing, cash flow after debt service each year, and the sale at exit (exit year NOI over the exit cap, less selling costs and the loan payoff).
Engine: buildWaterfall, lib/engine/waterfall.ts
Inputs: Purchase price, Closing costs, Loan to value, Origination fee, Interest rate, Amortization, Apartment units, Average monthly rent, Residential vacancy, Commercial area, SF, Commercial rent per SF, Commercial vacancy, Other income, Real estate taxes, Insurance, Management, share of EGI, Repairs and maintenance, Utilities, Marketing and G&A, Reserves, share of EGI, Rent growth, Expense growth, Other income growth, Hold, Exit cap rate, Selling costs, Loan term.
Every dollar distributed to equity divided by every dollar equity contributed, over the same cash flows as the IRR.
Engine: buildWaterfall, lib/engine/waterfall.ts
Inputs: Purchase price, Closing costs, Loan to value, Origination fee, Interest rate, Amortization, Apartment units, Average monthly rent, Residential vacancy, Commercial area, SF, Commercial rent per SF, Commercial vacancy, Other income, Real estate taxes, Insurance, Management, share of EGI, Repairs and maintenance, Utilities, Marketing and G&A, Reserves, share of EGI, Rent growth, Expense growth, Other income growth, Hold, Exit cap rate, Selling costs, Loan term.
PURSUE when project IRR is at least the 15.00% hurdle and min DSCR is at least the 1.00x debt coverage floor; otherwise PASS.
Engine: buildDecision, lib/engine/decision.ts
Inputs: Project IRR and Min DSCR above.
Worked example
The desk's sample deal, as it opens.
Used by: Going-in cap rate, Senior loan, Equity required, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Equity required, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 effective gross income, Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 effective gross income, Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 effective gross income, Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 effective gross income, Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 effective gross income, Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 effective gross income, Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 effective gross income, Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Year 1 NOI, Going-in cap rate, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Senior loan, Equity required, Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Equity required, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Min DSCR, Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Project IRR, Project MoIC.
Worked example
The desk's sample deal, as it opens.
Used by: Project IRR, Project MoIC.
Underwrite Core
Free public records for this address. A New York City lookup reads DOB permits, DOB violations, HPD violations, ACRIS deed and mortgage index rows, and the latest DOF assessment record. Another city returns one coverage row: not available for this city. The address goes to the Census geocoder, and a New York City match also goes to NYC Open Data. Nothing here changes rent, expenses, financing, or the verdict.
The worked example uses a saved sample. Live lookups require Underwrite Core.
Start your free monthSaved sample for the worked example. No live lookup was sent.
Public building records for this address: not available for this city.
Coverage
not available for this city
Source: Underwrite. Public records coverage. records-risk-coverage
Date:
Public data
Off until you include it for this deal. Then it lists free public records for this address: county and city assessor rolls, the LL97 covered buildings list, flood, hazard, environmental, market and rate context. Each fact shows its agency, dataset, record, fetch date and a link. Context only: nothing here changes rent, expenses, financing, the grade or the verdict.
Ask what would change if an assumption moved. The analyst proposes input edits and does not write a return, a coverage figure, or a verdict. Each card is the engine, run again, for that one input. Accept records who, when, the old value, the new value, and the rationale. Reject leaves the deal as it is.
On the worked example the answer is a recorded sample. On your own deal, Underwrite Core is required and the request is limited per person.
No accepted edits yet.
Underwrite Core
Nearby apartment and multifamily asking rents. A lookup sends the property address to the Census geocoder and its rounded coordinates to RentCast. Nothing enters the model until you choose Use in model.
Live local comps require Underwrite Core. The worked example uses a saved fictional sample.
Start your free monthAsking rent from active and recent listings. Not an appraisal.
Illustrative saved sample. These are fictional listings for the worked example.
Source: RentCast listing. Fetched . Search radius 1 mi. Up to 500 records searched per radius, with up to 120 usable listings shown. Coverage may be incomplete.
Deal average $2,400 is above market: +$725 (43.3%) versus matching-bedroom median $1,675. Middle range (25th to 75th percentile) $1,588 to $1,763, from 8 listings with 2 bedrooms.
Below or above market means outside the middle range. The deal average may include a different bedroom mix, concessions or utilities. Multiple listings can come from one building.
Scroll the table sideways to see all listing details.
| Address | Mi | Beds | Baths | SF | Rent / mo | $/SF | Status | Last seen |
|---|---|---|---|---|---|---|---|---|
| 100 Sample Street, Ithaca, NY 14850 · RentCast listing · fetched 2026-10-18 · sample | ||||||||
| 100 Sample Street, Apt 1, Ithaca, NY 14850Listed 2026-07-01 | 0.00 | 2 | 1 | 900 | $1,500 | $1.67 | Inact. | 2026-10-15 |
| 100 Sample Street, Apt 2, Ithaca, NY 14850Listed 2026-07-01 | 0.09 | 2 | 1 | 920 | $1,550 | $1.68 | Active | 2026-10-15 |
| 110 Sample Street, Ithaca, NY 14850 · RentCast listing · fetched 2026-10-18 · sample | ||||||||
| 110 Sample Street, Apt 1, Ithaca, NY 14850Listed 2026-07-01 | 0.17 | 2 | 1 | 940 | $1,600 | $1.70 | Active | 2026-10-15 |
| 110 Sample Street, Apt 2, Ithaca, NY 14850Listed 2026-07-01 | 0.26 | 2 | 1 | 960 | $1,650 | $1.72 | Inact. | 2026-10-15 |
| 120 Sample Street, Ithaca, NY 14850 · RentCast listing · fetched 2026-10-18 · sample | ||||||||
| 120 Sample Street, Apt 1, Ithaca, NY 14850Listed 2026-07-01 | 0.34 | 2 | 1 | 980 | $1,700 | $1.73 | Active | 2026-10-15 |
| 120 Sample Street, Apt 2, Ithaca, NY 14850Listed 2026-07-01 | 0.43 | 2 | 1 | 1,000 | $1,750 | $1.75 | Active | 2026-10-15 |
| 130 Sample Street, Ithaca, NY 14850 · RentCast listing · fetched 2026-10-18 · sample | ||||||||
| 130 Sample Street, Apt 1, Ithaca, NY 14850Listed 2026-07-01 | 0.52 | 2 | 1 | 1,020 | $1,800 | $1.76 | Inact. | 2026-10-15 |
| 130 Sample Street, Apt 2, Ithaca, NY 14850Listed 2026-07-01 | 0.60 | 2 | 1 | 1,040 | $1,850 | $1.78 | Active | 2026-10-15 |
| 140 Sample Street, Ithaca, NY 14850 · RentCast listing · fetched 2026-10-18 · sample | ||||||||
| 140 Sample Street, Apt 1, Ithaca, NY 14850Listed 2026-07-01 | 0.69 | 1 | 1 | 1,060 | $1,900 | $1.79 | Active | 2026-10-15 |
| 140 Sample Street, Apt 2, Ithaca, NY 14850Listed 2026-07-01 | 0.77 | 1 | 1 | 1,080 | $1,950 | $1.81 | Inact. | 2026-10-15 |
HUD SAFMR / FMR: unavailable for this lookup.
ACS median gross rent: unavailable for this lookup.
HUD payment-standard rents and ACS survey gross rents are public context. They are not current asking-rent comps.
Reading a document is part of Underwrite Core. Underwrite Core. $49 per user per month. 30 day trial. Renews monthly until canceled. Cancel anytime.
Upload a PDF, .pptx, or Excel file, or paste deal text. Review the extracted assumptions before using them.
Multifamily is the strong path. Mixed-use is supported when its leases are whole years. Model limitations.
Drop an offering memorandum PDF or an Excel pro forma, or paste text from a listing or memo. AI fills the inputs panel with the deal assumptions. A row shows its source sentence only when the server found it in your pasted or spreadsheet text and it states the value. Every other row carries an amber hand-verify flag with the reason, including every row from a PDF, whose text the server never has, and every tenant row. A value the server computes says how, and office or other space the model has no line for is named and left out. Extracted values enter the engine the same way typed ones do.
Privacy: when you click Parse, a PDF is sent to the processing service as a document; a .pptx is read into slide text first; an Excel file is read in your browser and only its cell text, normalized and capped in size, is sent. Files over 4 MB need private document storage, which is not turned on here. The processing service does not train on this input under its commercial API terms. Underwrite servers keep no copy. The server log records sizes and timings, and quotes the figure or word behind any parser note it withholds. Parsed values land in your browser. Do not upload confidential material unless you are authorized to send it.
Multifamily is the strong path. Mixed-use is supported when its leases are whole years. Model limitations.
Parse sends your source to the processing service. Upload only material you’re authorized to send. Review the privacy and model coverage details above.
Reading a document is part of Underwrite Core. Underwrite Core. $49 per user per month. 30 day trial. Renews monthly until canceled. Cancel anytime.
Read a rent comp or sale comp report you exported (PDF, Excel, CSV or pasted rows). Each row keeps the line it came from. Counts, ranges and medians are calculated from figures confirmed against the file, and set beside your assumptions. A PDF's rows are shown as read and are not confirmed against the file. A comp never changes the verdict or an input unless you use its median.
Upload only a comp file you have the right to use. Underwrite has no comp data of its own and reads only the comp file you give it.
Privacy: when you click Read comps, the file goes the way a deal document does. A PDF is sent to Anthropic as a document; an Excel or CSV file is read in your browser and only its cell text and file name are sent. Anthropic does not train on this input under its commercial API terms. Underwrite servers keep no copy; the server log records sizes and timings. The rows land in your browser, which keeps them for the memo until you remove them, load another deal, reload the desk, switch Confidential Mode on or clear site data.
Reading a document is part of Underwrite Core. Underwrite Core. $49 per user per month. 30 day trial. Renews monthly until canceled. Cancel anytime.
Upload floor plans as a PDF or an image. Underwrite reads the areas the plans print (gross, rentable, each unit and each floor) with the page each is on. It does not measure a drawing or scale a dimension. You review and edit every row, then pick the input it is for. Nothing reaches the deal until you click Accept.
Privacy: when you click Read plans, the file goes the way a deal document does in the parser above, under the same terms. Underwrite servers keep no copy. The rows stay in this panel until you clear them or load another deal.
No AI-extracted fields this session. The results come from the inputs you see.
The return is calculated from the inputs on screen. How the math works
Saved deals live in this browser's storage. Share and memo links encode the deal in the URL. Pasted source text is sent to Anthropic only when Parse is clicked.
Do not paste confidential material unless you are authorized to transmit it. What leaves your browser
Labels and the memo follow Mixed-Use. Expense recoveries are included only when you add them. Hotel furniture and equipment splits stay with you. Office tenant improvements and leasing costs follow each tenant on the rent roll.
Base-case inputs are active.