jevrecipes

Recipe catalog / expense-category

Match an expense to a category list

Does expense clearly fall under one of the caller's categories, under several equally, under none, or is it unclear?

You need to know whether a described expense can be filed under one of your own category definitions before an agent picks the category or asks the submitter for more detail.

Explore this recipe interactively ยท Source and implementation guide

Use expense-category in TypeScript

Install with npm install jev-recipes. Requires Node.js 22.9 or newer and ES modules. Set TYPESAFE_API_KEY in your server environment for live calls, which send input to TypeSafe and use API quota. See the installation guide.

import { expenseCategory } from 'jev-recipes/expense-category';

const result = await expenseCategory({
  "expense": "Uber from the airport to the client's office for the Tuesday kickoff meeting, then back to the hotel.",
  "categories": "Travel - Airfare: flights and airline fees.\nTravel - Ground: taxis, rideshares, rental cars, trains, and parking while traveling for business.\nMeals: food and drink while traveling or hosting clients.\nSoftware: subscriptions and licenses for work tools.\nOffice Supplies: consumable items for the office.",
  "minConfidence": 0.8
});
console.log(result);

Input contract

FieldTypeNeeded
expensestringRequired
categoriesstringRequired
minConfidencenumberOptional
Full input and result schemas
{
  "input": {
    "$schema": "https://json-schema.org/draft/2020-12/schema",
    "type": "object",
    "properties": {
      "expense": {
        "type": "string"
      },
      "categories": {
        "type": "string"
      },
      "minConfidence": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      }
    },
    "required": [
      "expense",
      "categories"
    ]
  },
  "result": {
    "$schema": "https://json-schema.org/draft/2020-12/schema",
    "type": "object",
    "properties": {
      "model": {
        "type": "string"
      },
      "usage": {
        "type": "object",
        "properties": {
          "input_tokens": {
            "type": "integer",
            "minimum": 0,
            "maximum": 9007199254740991
          },
          "output_tokens": {
            "type": "integer",
            "minimum": 0,
            "maximum": 9007199254740991
          }
        },
        "required": [
          "input_tokens",
          "output_tokens"
        ],
        "additionalProperties": false
      },
      "status": {
        "type": "string",
        "enum": [
          "ready",
          "review"
        ]
      },
      "verdict": {
        "type": "string",
        "enum": [
          "matched",
          "multiple",
          "none",
          "unclear"
        ]
      },
      "confidence": {
        "type": "number",
        "minimum": 0,
        "maximum": 1
      },
      "probabilities": {
        "type": "object",
        "propertyNames": {
          "type": "string",
          "enum": [
            "matched",
            "multiple",
            "none",
            "unclear"
          ]
        },
        "additionalProperties": {
          "type": "number",
          "minimum": 0,
          "maximum": 1
        },
        "required": [
          "matched",
          "multiple",
          "none",
          "unclear"
        ]
      }
    },
    "required": [
      "model",
      "usage",
      "status",
      "verdict",
      "confidence",
      "probabilities"
    ],
    "additionalProperties": false
  }
}

Saved example result

This hand-authored response demonstrates the contract. It is not a model accuracy measurement. Run it without an API key: npx jev-recipes demo expense-category.

{
  "model": "demo-fixture",
  "usage": {
    "input_tokens": 0,
    "output_tokens": 0
  },
  "status": "ready",
  "verdict": "matched",
  "confidence": 0.86,
  "probabilities": {
    "matched": 0.86,
    "multiple": 0.08,
    "none": 0.03,
    "unclear": 0.03
  }
}

Evaluation evidence

Fixture only

No verified live accuracy measurement is available. Evaluate representative cases before using this decision in your workflow.

Use the evaluation guide to measure this decision on your own labeled cases.

Limitations

Related recipes