Vendor bill review

Know what you owe.
Pay what you should.

Verlet checks vendor bills against your contracts and actual usage. Find charges to challenge, approve the claim and track the credit or refund.

Independent of the vendor, AI model or software you use.

Delivered
240
Circuits active for the full month

Contract. Usage. Invoice.Illustrative

01The equation

1 / 4

Every bill is an equation.

Charged equals agreed rate times delivered. The invoice is checked against the approved terms and delivery record.

When a vendor bills what was agreed for what actually happened, the two sides match.

01The equation

2 / 4

When they don’t,
something is left over.

$360,000 charged. $288,000 owed. $72,000 potential recovery, subject to review.

240 circuits · $1,500 billed · $1,200 agreed per month · Illustrative record, USD

01The equation

3 / 4

An equals sign with
something left over.

The top line represents what was agreed. The bottom line represents what was charged. The separate square represents potential recovery.

The mark is the comparison, reduced.

01The equation

4 / 4

A difference worth checking.

The comparison becomes the Verlet wordmark: two equal lines and a remainder.

A finding starts the review. An applied credit or received refund completes it.

Continue

03From finding to payment

Found is not recovered.

We bring the records together and check the charges. You approve each claim before it goes to the vendor. Only an applied credit or received refund counts as recovered.

Contracts and amendments
Usage and service records
Invoices and credits
VerletCheck the charge
Amount owed
Claim for approval
Credit or refund
Each finding includes the contract terms, usage and calculation.
  1. 01

    Finding

    The difference and its calculation.

  2. 02

    Human review

    You approve the claim.

  3. 03

    Vendor response

    The vendor accepts or disputes it.

  4. 04

    Credit applied

    Credit applied or refund received.

04The team

Work with the founders.

Mike Lunati

Led data operations at Scale AI for six years. Previously McKinsey, Morgan Stanley and Hg.

Liza Fireman

Built security software at Datadog. Led engineering at Snowflake, Meta and Google.

Our team built at
  • Scale AI logo
  • Datadog logo
  • Snowflake logo
  • Meta
  • Google
  • McKinsey & Company
  • Morgan Stanley
  • Hg

05Start here

Which bill should we check?

Tell us the vendor and billing period. We will agree what to review and how to share the files.

Check a bill