Expense category matching: six recorded examples
The downloads contain six original inputs, their expected labels and one recorded API observation per input revision. Disagreements stay visible.
Supply purchased items and their business purpose to compare software, travel, office and meal categories. Keep mixed or unsupported purchases for review rather than deciding tax treatment.
The taxonomy is intentionally small. A receipt can be authentic and still fall outside the four named expense categories. Review covers mixed receipts and missing purchase descriptions instead of guessing from a merchant.
The demo returns a proposed internal category. Amount parsing, currency conversion, accounting rules and ledger writes remain outside the model request. No tax or financial treatment follows from the label.
This is an independently authored text demonstration inspired by the linked X discussion. Its rules and synthetic inputs are ours. It does not reproduce the author’s application, verify reported outcomes, or perform an external action.

Use the sample to understand the rule, then test your own input for expense category matching.

Supply purchased items and their business purpose to compare software, travel, office and meal categories.

Open the decision rules and keep the intended distinction between Software, Travel, Office, Meals, Review. Edit the choices only when testing your own rule.

Use a sample first, then try your own input. Compare the selected label with the other returned probabilities.

Review Review before using a result. The label is a decision about the supplied information, not an executed action.
| Original input | Observed label | Leading probability |
|---|---|---|
| Hosted database | Software | 100% |
| Train to meeting | Travel | 100% |
| Printer paper | Office | 100% |
| Team lunch | Meals | 100% |
| Unknown order | Review | 100% |
| Mixed receipt | Review | 83.0% |
These are six original synthetic inputs, tested once per request revision. The percentage is the leading option probability, not measured accuracy. Authored expectations and any disagreements remain in the downloadable evidence. 6 of 6 results matched the authored fixture expectations.Verified
The downloads contain six original inputs, their expected labels and one recorded API observation per input revision. Disagreements stay visible.
The single-request SDK runner uses your private TYPESAFE_API_KEY. The offline consumer reads saved observations, checks request hashes and prints a pending output. It does not call the model or change any external system.
# Offline: inspect the downloaded saved observations. No API key is needed.
python preview_choice_workflow.py samples.json saved-api-results.json
# Optional live request: install the SDK and set TYPESAFE_API_KEY privately.
npm install @typesafe-ai/sdk@0.6.0
node run-choice-sample.mjs samples.json case-1
# The live response prints to the terminal; it does not update saved-api-results.json.The input contract and these limits define what the result means.
This example returns Review. A separate line-item workflow could classify each item, but the current input receives one label.
Yes. Update the rule and choice list together, then replay representative receipts. A label that lacks a clear definition can make the result harder to interpret.
It describes the distribution over the supplied choices for this request. These six synthetic examples are functional tests with authored expectations, not a production accuracy benchmark.
No. The demo displays a label and distribution. The downloadable consumer produces a pending preview packet. Application code and a separate authorization policy would control any real action.
Edit the example above to see how a changed fact affects the decision, or open the Playground for another task.