Data Normalization

Data processing and normalization tasks evaluate structured output from messy prose and different structures. This capability is essential for catalogue and product pipelines where data needs to be extracted from unstructured text and formatted consistently. Models must understand the desired output format and extract relevant information accurately.

Duration

Input Tokens

1015

Output Tokens

280

Cost

$0.00

Context

Input

A customer from the USA is interested in the following product, return the product sheet. Base the response only in the context.

Expected output

{
  "country": "USA",
  "product": "AeroSeal-X Polymer Matrix",
  "ingredients": [
    "NanoSilicon Oxide-Δ3",
    "Isomerized Hexaoxy Toluate (IHT)",
    "Zinc Pyrithionate Modulator",
    "Tetramethyl Ethyloxy Resinate",
    "Polyether Vinylcaprolactam",
    "ThermoReactive Benzonate (TRB-9)"
  ],
  "allowed": false
}

Model output

{
  "country": "USA",
  "product": "AeroSeal-X Polymer Matrix",
  "ingredients": [
    "NanoSilicon Oxide-μ3",
    "Isomerized Hexaoxy Toluate (IHT)",
    "Zinc Pyrithionate Modulator",
    "Tetramethyl Ethyloxy Resinate",
    "Polyether Vinylcaprolactam",
    "ThermoReactive Benzonate (TRB-9)"
  ],
  "allowed": false
}