Skip to main content
Menu

What we do

Raise CapitalThe investors who fit, and meetings with them.Source DealsEvery company in your niche, and who owns it.Grow RevenueWe find customers and partners, and who can introduce you.Map the MarketKnow your market before you commit to it.Underwrite and DiligenceWe check a deal's claims before you sign.Partner firmsFor firms running a raise or a search for their own clients.Data ServicesCustom data from public records, kept current.

The firm

ResearchAboutHow We WorkPlatformInsightsSchedule a conversation

Commercial due diligence and acquisition underwriting

Before you sign, bid or lend, we test the deal: the documents, the business and the numbers.

For private equity firms, sponsors, family offices, lenders and credit funds, and the companies and funds they back.

Schedule a conversationSend us one deal or asset

Due diligence and underwriting: four practices, one standard

Is it true?

Will it pay?

  1. At the teaser or memorandum

    What the deal says

    Every claim in the pitch, checked.

  2. When the data room opens

    The whole data room

    Every file, read and tied together.

  3. Before a binding bid

    Commercial due diligence

    Whether the business case holds.

  4. Before you bid or lend

    Underwriting

    Property, loans, private credit and SBA loans.

One standard under all four

Every practice on one standardThe practices, what the deal says, the whole data room, commercial due diligence and underwriting, stand on one standard: each claim on its own line, checked against public records, numbers rebuilt and hard cases run, where they disagree, a second review that tries to disprove it, and your decision.

  1. Each claim on its own line

  2. Checked against public records

  3. Numbers rebuilt, hard cases run

  4. Where they disagree

  5. A second review tries to disprove it

  6. You decide

Start at any column.

What the deal says, checked before you sign

We read
The deck, memorandum, CIM, model and fund documents, checked against filings, court dockets, liens and permits.
You get
A plain answer on each claim, the model rebuilt and run on the cases it left out, and who is on the other side.

Real work

We checked one fund manager as an investor would. A second review then tried to disprove each finding: 77 held up, 8 were corrected and 3 were removed.

Delivered

One sponsor's offering memorandum to its own investors showed roughly 5x. We rebuilt its model within 0.6% of its base case; with flat earnings, investors would make about 1.9x.

What a second review did to each finding: most held up, a few were corrected, a few removedSmall squares in rows, one for each finding. The rows in ink held up. In the last row, a run in a second colour was corrected, and the last few, struck through, were removed.Held upCorrectedRemoved

The data room, and everything you are handed

We read
The model, monthly accounts, loan agreements, cap table, customer records and any research you were given.
You get
Where the documents disagree, what is missing, and a document request for each gap.

Delivered

With no data room yet, the research behind one deal ran to about 45,000 words in six documents. Only two were about the company, and they disagreed on what it sells.

Real work

In one full data room, we read all 42 files. Earnings were stated three ways and the raise size five, and there was no audit, though a loan covenant required one.

A research pack: mostly background, and the parts about the company disagreeA row of document sheets. Most, drawn quietly, are background. The last pair, drawn in ink, are about the company, and a line on each is marked and tied to the other: they disagree.A research packDisagreeBackgroundAbout the company

Commercial due diligence

We read
Government registers, city contracts, permits, court records, published prices, customer reviews and what the seller shares.
You get
The case against the deal up front, a request list for the seller, and a report and live model with every input sourced.

Sample

From only a codename and a short description, we found the company, sized its market, mapped its customers, competitors and prices, and tested the buyer's thesis. It ended in three conditions to meet before a binding bid, with 2,295 sourced claims behind them.

From a codename to an answer, through the market, the customers, the competitors and the thesisA codename card leads through public-record stops to linked boxes that turn out to be one company. From there the work splits into the market, the customers, the competitors and the buyer's thesis, and each feeds a field of small marks, each standing for a set number of sourced claims, which runs through the conditions, drawn as gates, to a filled square: the answer.A codenamePublic recordsOne companyMarketCustomersCompetitorsThesisEach mark is fifteen claimsConditionsThe answer

Underwriting property, SBA loans and private credit

We work out what the numbers support, run the hard cases and show where the deal breaks.

  1. Properties, small to large, each with a price rangeA row of building outlines growing from a small office to a large apartment complex. Under each one, a short line with end marks shows the range of prices that still works.Price rangeOfficeApartments

    Before a bid

    Property

    The price range that still earns your target return, and what would sink the deal.

    Real deals

    Nine property deals, from a $1MM office to a $50MM apartment complex.

  2. Public records pulled on one loan, and the one that showed a flood zoneA dense block of tiny document marks, the records we pulled. One document is drawn out of it, large, with a hatched band marking the flood zone.Records we pulledFlood zoneOne of them

    A loan you are making, buying or holding

    Loans and private credit

    A new loan: what the borrower and the collateral can carry, and where it breaks. One you hold: what the lender could recover, from servicer, county, court and lien records.

    Pilot

    628 public records pulled across our research lanes on one distressed apartment loan. One put the property in a flood zone the client's team hadn't uncovered.

  3. Loans pulled from a book, each opened into its own fileA field of small loan marks. A few are pulled out in ink, and each opens into a file card of ticked checks: liens, courts, state filings, exclusions and similar loans.LiensCourtsState filingsExclusionsSimilar loans

    One loan at a time

    SBA loans

    Each borrower's liens, court cases, state filings and federal exclusions, and how comparable SBA loans performed.

    Sample

    Three sample loans from a credit manager, each with its own credit memo and a source beside every finding.

We underwrite and report. The lender or investor decides.

Real work for clients and prospects, names withheld.

Sources and dates
Sources and dates for the figures on this page
FigureSourceAs of
77 held up, 8 corrected, 3 removedOur written review of one fund manager, read as an investor weighing an allocation would read it; two adversarial review passes then tried to disprove each finding. Praxis research file.August 2026
Roughly 5x, within 0.6%, about 1.9xOur written review of the offering memorandum and model one sponsor put to its own investors for an acquisition under LOI. The memorandum showed 5.1x to investors in its base case; our returns engine matched that base case within 0.6%, and with flat earnings it returned about 1.9x. Delivered with the review. Praxis engagement file.August 2026
About 45,000 words in six documentsThe research pack the client sent us with one offering memorandum under LOI, read without data room access in our written review, delivered to the client. Praxis engagement file.August 2026
42 filesOne company's full data room, read file by file into a claim register: one year's earnings stated three ways and the raise size five ways across its materials, and no audited statements, though its loan agreements require audited annual accounts. Praxis engagement file.September 2026
2,295 sourced claims and three conditionsA commercial due diligence report built from public records as a sample for a prospective client. Praxis sample file.September 2026
Nine property deals, $1MM to $50MMOur underwriting packages on nine real property deals, one package per deal. Praxis engagement files.August to October 2026
628 public records and the flood zoneRaw captures of public-record documents our team pulled across research lanes on one distressed apartment loan in a pilot; the flood zone is from the client's own reply. Praxis pilot file.August 2026
Three sample loansPer-loan credit memos on three sample loans a credit manager sent us, returned as a sample. Praxis sample file.September 2026
DiagramsDiagrams drawn from the figures above, not data.

What you get back

A memo that leads with the answer

Our answer first, then the evidence behind it.

A live spreadsheet

The model rebuilt, every input sourced.

A web edition

The whole package as a page you can explore.

A source ledger

Every claim with its source and exact quote.

Document requests

One request for each gap, tied to its finding.

An interview guide

Who to call, in what order, and what to ask.

Send us one deal or asset: a company, a property or a loan. We'll show you what checks out and what doesn't.

Schedule a conversationSend us one deal or asset

Not an appraisal or an opinion of value.

Choosing a market first? Map the MarketQuestions about fees and dataWho we're not forHow we trace every numberResearch