Skip to content

Industry field guide

Dealer-arranged finance contract data for AI testing

A dealership worksheet may describe a proposed deal while a finance company’s acceptance refers to a later contract.

2022 NAICS
522220
Updated

What this guide tests

Which D-A document has a matched acceptance, and does the second packet establish acceptance for D-B/C1?

See the specimen and fields · Work through the fictional records

Task and source context

A dealership worksheet may describe a proposed deal while a finance company’s acceptance refers to a later contract. CFPB’s auto finance pilot and consumer materials distinguish relevant parts of the dealer-arranged financing process. An evaluation can test document lineage using invented parties and amounts. CFPB auto finance data pilot · CFPB financing routes

Prepare a practice sale packet with a preliminary worksheet, a revised installment contract, and an acceptance record tied to the revised contract. Give a second case an acceptance notice from an earlier deal with a similar reference. Ask which contract can be matched and what evidence remains missing. Include a trade-in figure only as a document-matching field.

Leave real applications, signatures, lender buy rates, vehicle IDs, and credit decisions out of the sample. Clear dealership and finance-company rights separately before discussing production records.

Describe a sales financing dataset

Industry research

The U.S. Census Bureau’s industry definitions establish which activities these codes cover. The workflow sources in this guide describe the task or record type. Fiund’s pilot below uses fictional material to test a specific decision within that scope.

522220 · Sales Financing

This industry comprises establishments primarily engaged in sales financing or sales financing in combination with leasing. Sales financing establishments are primarily engaged in lending money for the purpose of providing collateralized goods through a contractual installment sales agreement, either directly from or through arrangements with dealers.

U.S. Census Bureau, 2022 NAICS Manual, p. 424 ↗

Build a pilot sample

Set up this pilot

Create two neutral contract revisions and an acceptance tied to the earlier one. Reuse the deal alias and score the actual accepted document without inheriting status.

Prepare the test packets

  1. Start with the worked case. Use the fictional record types below as a guide to the source files you need. Give each packet a stable ID and save its source versions with it.
  2. Change one fact. Make one packet with enough evidence to answer the question and one where the evidence points to a different answer. Keep the other fields fixed so a reviewer can tell what changed.
  3. Remove the decisive evidence. Give the reviewer a third packet with a plausible near-match in its place. Have them write what can still be established and what evidence would resolve the rest.

Fields the reviewer needs

Deal alias and document class
Separates proposals from contracts.
Contract revision and acceptance target
Establishes the exact match.
Practice figures and unresolved relationship
Exposes mismatched versions without guessing.

Use How to prepare a financial workflow data sample for AI to turn the specimen into a brief with a defined task, source and review owner.

Fictional exercise

Finance acceptance attaches to a contract revision, not a preliminary worksheet

A practice sale packet has invented deal aliases and arbitrary figures. It contains no signatures, credit applications, vehicle IDs or actual lender terms.

Records supplied to the evaluator
Record IDSourceWhat it records
DF1Worksheet W1Deal D-A, preliminary version: practice total 200 and trade-in field 20. Marked proposal only.
DF2Contract C2 and acceptance A2Revised D-A/C2: practice total 190, trade-in 25. A2 explicitly accepts D-A/C2, not W1.
DF3Second packetDeal D-B/C1 has an acceptance referring to older D-B0/C1. No D-B-to-D-B0 revision relationship supplied.

Question for the model. Which D-A document has a matched acceptance, and does the second packet establish acceptance for D-B/C1?

Show the reference answer

DF2 explicitly matches A2 to D-A/C2. DF1's proposal has different figures and is not the accepted document named in A2. DF3's acceptance points to D-B0, while the packet asks about D-B; the similar contract suffix does not establish a revision relationship. Flag the second packet's deal-identity gap rather than transferring acceptance. The task identifies documentary links without assessing finance suitability or a consumer decision.

Download the fictional records and answer key (JSON)

How to check the result

Ask for the answer, the record IDs that support it and any fact the evidence leaves unresolved. Have a reviewer compare the response against the reference before deciding whether the sample is ready to expand.

Errors to flag in this case

  • Summarizing accepted figures from the earlier proposal instead of the named contract.
  • Matching by a contract suffix while ignoring a conflicting deal reference.

Change the inputs and run it again

Add an explicit D-B0-to-D-B conversion record that excludes acceptance carryover. Require the identity link to be recognized while acceptance remains unconfirmed.

Score the decision and the evidence separately

Save the reviewer’s expected answer, supporting record IDs and unresolved fields in a separate answer key. Score the model on the decision, its citations and its handling of the deliberate gap. A correct answer citing a different case counts as an evidence error.

Before expanding collection, have a second reviewer run the same checks on the held-out packet. Keep related versions, views and exports of the same case together so they stay on one side of the training and evaluation split.

Check permission and scope

Use independently written fictional packets. Exclude credit decisions, lender buy rates, applications, signatures and actual dealership records.

Exclude real personal and financial data, credentials, underwriting decisions and individual eligibility conclusions.

Describe an authorized archive through Fiund’s existing-data assessment, or describe a new capture capability through the collection brief. The contributor agreement covers the terms for participation.

Sources and further reading

Resources for this workflow

How the examples work

Company and program reviews

Use these reviews to compare sourcing, capture or evaluation approaches. Confirm the applicable program, accepted sources and terms with the provider.

All company reviews

Related industry examples

← Finance and insurance