Product Tour

Inside Minervian AI: CRE underwriting, screen by screen

Every image on this page is a real screen from the live product, captured on office, multifamily, and development deals, with tenant and property names changed. Work down from the lease inputs and cash flow through financing, the equity waterfall, and the deal summary, or jump to the section you care about. Click any screenshot to enlarge it.

01

Lease-by-lease cash flow

Inputs › Leases › Recovery

Recovery terms set tenant by tenant, down to the base year

Every lease carries its own recovery structure. This illustrative tenant is on a gross lease with a base-year stop, so it pays its pro-rata share of operating expenses above the base year, with the OpEx share, admin fee, base-year OpEx, and annual cap all set on the lease itself. The AI Advisor on the left can change any assumptions across any leases in one instruction.

Inputs › Leases › Leasing Costs

Leasing costs resolved lease by lease

Tenant improvements, landlord work, and inside- and outside-broker commissions are set on each lease as a percentage of lease, a total, per square foot, or tiered, each with its own timing. One plain-language instruction can change a tenant’s commissions across all of its leases at once (2 changes here). The cash flow by tenant underneath shows the tenant improvements and leasing commissions for each lease.

Outputs › CF by Tenant (Commercial)

Monthly cash flow, tenant by tenant

The projection is built lease by lease and shown the same way: 23 tenants here, each on its own row, with expense recoveries calculated from that tenant’s own lease terms rather than a flat percentage. Vacant space runs as lease-up tranches that appear as their own rows, and every view switches between monthly and annual.

Outputs › Rent Roll

Several buildings in one deal

A deal holds as many buildings as it needs, and the rent roll breaks out by building. Here there are two, each with its own tenants, percent of building, weighted-average remaining term, and in-place rent, and a selector filters the view.

02

AI that fills in what the OM leaves out

AI Advisor › Auto-complete

Auto-complete proposes what the OM leaves out

Offering memoranda rarely supply a full set of assumptions, so Auto-complete lists what is missing, groups it by how much each item matters, and proposes values with a rationale for each. It can also resolve fields that contradict one another, such as dates outside the projection window, and you choose whether it writes to a new version first or fills the current one in place.

03

Multifamily value-add, by cohort

Inputs › Multifamily › Mark-to-Market · Outputs › CF by Cohort (MF)

Value-add bands instead of unit-by-unit guesses

Units are grouped into bands by how far below market they sit, and each band has its own turnover pace, months offline, capex per unit, and cash-for-keys budget. The program runs on those inputs rather than a date for every apartment. Cash flow by cohort underneath shows loss to lease falling as units move from pre- to post-renovation.

Inputs › Multifamily › Unit Summary · Outputs › CF by Cohort (MF)

The cost of the renovation program, month by month

The unit summary rolls the rent roll up by unit type, so assumptions are controlled at unit-type level. Below it, transition vacancy shows what the program costs: the units offline for renovation and the rent forgone while they are, month by month and band by band.

04

Multi-tranche financing

Outputs › Cash Flow (annual)

Land, construction, and permanent debt in one stack

Each tranche runs its own schedule; in this case: a land loan, a floating-rate construction loan drawn over two years on a forward SOFR curve plus spread, and a permanent loan that refinances the construction balance. Draws, fees, rate build-up, interest, and repayment are visible for every tranche and flow into the levered cash flow. For how other tools handle this, see the loan modelling comparison.

05

Equity waterfall

Outputs › Deal Summary › Waterfall

An equity waterfall described in plain language

Describe the structure the way you would brief an associate and the AI builds it: classes of securities rather than a fixed GP/LP template, a funding split, and tiers for return of capital, an IRR hurdle, a fixed-dollar bonus, and a profit-share test. The schedule shows what each tier paid to each class, and returns roll up by class and by investor. See how other tools compare in the equity waterfall comparison.

06

Deal summary and returns

Outputs › Deal Summary

Sources and uses, profit, and returns

The deal summary lays the model out as investment-memo-ready tables. This screenshot shows sources and uses, footing to the same total with per-square-foot and per-unit columns, as well as project returns including unlevered and levered profit, equity multiple, and IRR. These tables export to Excel for the memo.

Outputs › Deal Summary › Yield on Cost

Yield on cost, untrended and trended

Yield on cost is struck two ways: untrended, in today’s dollars at 93% occupancy, and trended, from the projection’s own stabilized year. The revenue-to-NOI build sits beside each, line by line.

Underwrite your next deal in minutes.

Sign up free and bring a real deal.