Skip to content
Get started

Internal expense-policy review

Match expense claims to explicitly approved internal rules, then review deterministic limits, missing receipt references and possible duplicates.

Input

Expense claims: id,date,claimant,merchant,description,amount,receipt_id E1,2027-03-02,TEAM-A,City Hotel,Overnight hotel stay for a client visit,140.00,REC-101 E2,2027-03-02,TEAM-A,Rail Company,Train journey for the client visit,35.00, E3…

Result

Both EUR 140.00 hotel claims exceed the explicit EUR 120.00 hotel limit and share receipt reference REC-101, so both remain visible with duplicate findings. The EUR 35.00 rail claim has no receipt reference. The claimed total remains EUR…

Review expense claims against supplied internal policy without inventing reimbursement rights or tax rules. The AI proposes semantic links between each claim description and approved rule IDs, backed by exact quotations. Server-side checks apply the explicit numeric limits, inspect receipt-reference presence and flag possible duplicates. Every original claim remains visible and included in the claimed total.

Your texts, pages and images are processed on our own servers in the EU. Trust Center

Try it now Pricing

Who it is for

Operations teams, finance administrators and small businesses preparing expense reviews from maintained internal rules. A responsible person confirms the supplied rules and checks the completed report before any reimbursement decision.

Typical uses

  • Review hotel, travel and other expense descriptions against approved rule statements.
  • Compare individual claims or daily claimant-and-rule groups with explicit limits.
  • Flag missing receipt references and repeated claims without deleting evidence.
  • Keep unmatched and ambiguous claims open for a human decision.

Example

Input

Expense claims: id,date,claimant,merchant,description,amount,receipt_id E1,2027-03-02,TEAM-A,City Hotel,Overnight hotel stay for a client visit,140.00,REC-101 E2,2027-03-02,TEAM-A,Rail Company,Train journey for the client visit,35.00, E3,2027-03-02,TEAM-A,City Hotel,Overnight hotel stay for a client visit,140.00,REC-101 Internal expense rules: id,description,max_amount,receipt_required R1,Hotel accommodation for business trips is reviewed against the supplied limit.,120.00,yes R2,Rail travel for business trips requires a receipt reference.,80.00,yes Confirm internal rules: These exact rules are approved Currency: EUR Limit scope: Per item Receipt requirement: Use approved rules Duplicate review: Both checks Review from date: 2027-03-01 Review through date: 2027-03-31 Review row order: Findings first Show evidence columns: Include quotations

Result

Both EUR 140.00 hotel claims exceed the explicit EUR 120.00 hotel limit and share receipt reference REC-101, so both remain visible with duplicate findings. The EUR 35.00 rail claim has no receipt reference. The claimed total remains EUR 315.00; no amount is approved for reimbursement.

A fixed example, shortened. Real results are AI-generated: please check them.

Price and limits

Price
usually 1–3 credits
Available with
Free, With package
Input limit
One page with up to 40,000 characters of HTML

Privacy

Processed on our own servers in the EU (Poland). Inputs and results are deleted after 7 days (Free) or 30 days (packages), or right after you see the result if you choose not to store them. Images are deleted right after processing.

In the API

Tool name: expense_policy_check. Create a job with POST /api/v1/jobs and fetch the result with GET /api/v1/jobs/{id}. You find your API key in your account.

curl -X POST https://do4u.eu/api/v1/jobs \
  -H "Authorization: Bearer d4u_…" -H "Content-Type: application/json" \
  -d '{"tool": "expense_policy_check", "url": "https://example.com/"}'

Full API documentation →

Questions and answers

Who approves the policy?

The responsible reviewer explicitly confirms the exact supplied rules, including their descriptions, limits and receipt requirements. Neither a claim description nor a model reply can grant that approval.

Does a matching rule approve reimbursement?

No. A match is a model proposal requiring human review. No matching rule, multiple plausible rules, missing receipts and excess limits remain visible findings. Even a row without a numeric finding is not approved.

What are the limits?

Up to 30 claims and 12 explicitly approved rules; JSON input up to 40,000 UTF-8 bytes and each CSV up to 24,000 bytes. Expense descriptions are limited to 600 bytes and rule descriptions to 800 bytes. Positive claim amounts and nonnegative rule limits cannot exceed 10000000.00. One currency with two decimal places; dates from 2000 through 2100. Each model proposal must reference exact existing source and rule IDs with a verbatim expense excerpt. Unknown or multiple rule matches remain unresolved. No receipt image verification, tax determination, reimbursement approval or payment is performed.

Is my text stored?

Only as long as needed: 7 days on Free, 30 days with a package – or not at all if you choose not to store it. Processing happens on our own servers in the EU (Poland).

Can I use the result right away?

Results are AI-generated. Please read them before you publish or send them.