Quickstart

Push a borrower and their documents, then read back a memo.

Push the borrower, loan terms and any files in a single call, then poll until the documents are extracted.

curl

curl -X POST https://underwriter.kita.ai/api/v1/intake \
-H "Authorization: ApiKey $KITA_UW_API_KEY" \
-F 'application={"business_name":"Delta Fabrication LLC","borrower_email":"owner@delta.example","loan_type":"SBA 7(a)","loan_amount":250000,"external_ref":"LOS-88213"};type=application/json' \
-F "files=@2024-tax-return.pdf" \
-F "files=@bank-statement-jan.pdf"

Then poll for extraction

Documents move from awaiting to processing to verified or low_confidence. Poll the document list until nothing is still in flight.

import time
 
while True:
    docs = requests.get(
        f"{BASE}/applications/{app['id']}/documents", headers=headers
    ).json()["data"]
    if all(d["status"] not in ("awaiting", "processing") for d in docs):
        break
    time.sleep(5)

Then read the credit picture and memo

credit = requests.get(f"{BASE}/applications/{app['id']}/credit", headers=headers).json()["data"]
print(credit["spread"], credit["decision"])
 
requests.post(f"{BASE}/applications/{app['id']}/memo", headers=headers)
memo = requests.get(f"{BASE}/applications/{app['id']}/memo", headers=headers).json()["data"]

Next: the end-to-end workflow in full.

On this page

Then poll for extraction Then read the credit picture and memo