Underwriting data extraction

Step 01

Define and Align

Upload sample data, define your output schema, and provide feedback on the hardest cases.

Output schema

Risk locationText · optional: Location of the insured risk, distinct from the mailing address.
AreaNumber · optional: Stated building area, paired with its source unit.
UnitChoose one · optional: The unit attached to the area in the source.
sq_ft
sq_m
null
SprinkleredYes / No / Unknown: Whether sprinkler coverage is explicitly documented.
true
false
null

app.sutro.sh/functions/underwriting-data-extraction

Collecting feedback · 1/3

Extract supplied risk attributes from synthetic underwriting submissions.

Example 1 of 3

Different locations

Model input

Document excerpt

Underwriting submission

Synthetic submission: mailing address in Portland; insured warehouse in Salem.

Model output

Risk location

Area

Unit

Sprinklered

Why? Optional guidance

Step 02

Model Sweep

Compare quality and cost on your examples, then choose a model or use Sutro’s default.

Compare models on expert agreement and cost. The interactive chart uses illustrative measurements.

Illustrative sweep · select Value or Best to use that model in your function.

Step 03

Deployment

Call Sutro Functions like any other software. Swap the model as soon as something more performant is released.

Selected modelWaiting for sweep results.

Real-time

import sutro as so

input_data = {
    "text": "Synthetic submission: mailing address in Portland; insured warehouse in Salem."
}

result = so.run_function(
    "underwriting-data-extraction",
    input_data,
)

curl --fail-with-body -X POST "${SUTRO_API_URL%/}/v1/functions/underwriting-data-extraction/run" \
  -H "Authorization: Key $SUTRO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "input": {
    "text": "Synthetic submission: mailing address in Portland; insured warehouse in Salem."
  }
}'

Batch

import sutro as so

job_id = so.batch_run_function(
    name="underwriting-data-extraction",
    data="underwriting_data_extraction.csv",
)
results = so.await_job_completion(job_id)

curl --fail-with-body -X POST "${SUTRO_API_URL%/}/v1/batch-inference" \
  -H "Authorization: Key $SUTRO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "underwriting-data-extraction",
  "inputs": [
    {
      "text": "Synthetic submission: mailing address in Portland; insured warehouse in Salem."
    }
  ]
}'