An AI credit analyst for small-ticket equipment finance.

Stop reconstructing borrower cash flow by hand.

Upload the deal package. Get a normalized cash-flow summary with every number cited back. Minutes, not days.

Book a 15-minute demo See a sample deal

Built by an engineer who has done this before. Pilot program open. No production data required.

Pinewood Lumber LLC Skid steer — $112,400 · 60 mo
In review
  1. Three-year trailing revenue $1,284,310 TR ’21–’23 · p. 4
  2. Bank-deposit cross-check $1,302,748 BS 12mo · export
  3. Owner-draw add-back + $42,000 P&L ’23 · L. 14
  4. Depreciation add-back + $38,910 TR ’23 · Sch. C
  5. Seasonal-variability haircut − $7,250 BS · Q1 deposits
  6. Qualifying cash flow $158,420 computed
Every number cites back to a source.

Your team is doing this in Excel and email.

Today, your credit officers receive a deal package — tax returns, bank statements, P&L, equipment invoice — and rebuild the borrower’s qualifying cash flow by hand. They normalize for owner draws and depreciation. They scan bank statements for seasonal swings. They cross-check vendor invoices against industry benchmarks. They write conditions one at a time in an email. They do this five to fifteen times a day.

On every deal, the math is correct or it isn’t. The audit trail is defensible or it isn’t. The decision is faster than your competitor or it isn’t. Right now, all three depend on which analyst opened the file and how tired they were.

Upload the package. Get the rebuild. Stay in the seat.

  1. You upload the deal package.

    Bank statements, tax returns, P&L, application, equipment invoice. PDFs from email, scans from the vendor portal, exports from QuickBooks. Whatever your team already has. No new intake form. No integration.

    • PDF
    • Scan
    • QBO export
    • Email forward

  2. You get normalized cash flow, with citations.

    Three-year trailing average. Bank-deposit cross-check. Seasonal-variability haircut. Owner-draw and depreciation add-backs. Every line cites the source document and page. DSCR computed against your threshold. Confidence score on the document set.

    • Trailing average
    • Deposit cross-check
    • Add-backs
    • DSCR

  3. You get draft conditions, and a chat over the file.

    A first-draft conditions list your analyst can edit and send. A chat panel that answers “why is DSCR 1.78?” and “is the VIN clean?” with citations, not vibes. Your analyst stays in the seat — the AI does the recap.

    • Draft conditions
    • Chat over file
    • Cited answers
    • Editable

Built for credit officers, not for a demo.

Every number cites back.
No black box. If the AI says $158,420 qualifying cash flow, click any line to see the source document and page. Your audit trail writes itself.
You stay in the seat.
This is a workbench, not an auto-decisioner. The AI proposes; the analyst approves, overrides, or adds. Your existing credit policy doesn’t change.
Designed for small-ticket reality.
Equipment finance isn’t mortgage. We don’t ship a mortgage tool with the word equipment in the marketing. The patterns, the doc types, the seasonal logic, the vendor-fraud heuristics — built for your deals.

Made for the analyst on file #14 of the day.

If you’re a credit officer, underwriting ops manager, or operations lead at an independent equipment-finance lessor, a captive, or a bank-affiliated leasing arm — and your team handles small-ticket deals between $25,000 and $250,000 on legacy LMS systems with Excel and email patched around them — we built this for you.

We’re not building a Loan Origination System. We’re not selling to bank IT. We’re a focused workbench that sits alongside what you already have and removes the most painful thirty minutes per file.

Pilot program

We’re working with a small group of design partners. Want in?

We’re not in general availability. We’re running a discovery sprint with a handful of equipment-finance lenders to make sure we build exactly what your credit team needs — not what we think they need.

Design partners get

  • A 15-minute demo on a redacted sample deal — or a redacted version of one of yours, once we sign a mutual NDA.
  • Free access to the prototype during the pilot.
  • Direct line to the engineer building this. Your feedback shapes the roadmap.
  • First crack at production pricing when we GA.

In exchange we ask for

  • Thirty minutes a week.
  • Honesty about what’s broken.
  • One redacted sample deal we can use to demo your workflow back to you.

Five fields, no more. We respond within two business days.

Built by someone who has done this before.

I spent the last few years as an early engineer at a mortgage AI underwriting platform — built the variable-income calculation system, the document cross-check, the chat-over-conditions. The features that pulled hardest with real underwriters were the ones that did the math correctly with a defensible audit trail. That insight is the entire premise of this product, now built clean from scratch for equipment finance. Bootstrapped. Focused. Talking to lenders, not investors.

— Founder, Tally Hand