Amazon Textract for automatic signature identification

Overview

Development stage

In development or acquisitionSource: a) Pre-deployment – The use case is in a development or acquisition status.

High-impact designation

Not high-impactSource: c) Not high-impact

Impact justification

Not reported

Start date

2025-05-01 (day precision)Source: 2025-05-01T00:00:00

Withheld from public reporting

NoSource: a) No

Mission

Topic area

Administrative Functions

Operational functions

No function tag matched Derived by keyword rules; see methodology

Problem the AI is intended to solve

The Homeownership Center (HOC) staff is required to review and validate each Case Binder.

Expected benefits

Amazon Textract automatically analyzes PDF documents to detect if the document has a signature panel and if the document is signed or not. This tool will help to automate and improve the overall efficiency of the case binder business process within Federal Housing Administration (FHA) Catalyst and potentially save about 30 minutes time.

System outputs

Determination of whether or not case binder signature panels have been signed.

Technology

AI classification

Natural language processingSource: Natural Language Processing: AI that processes, interprets, and shares information in human language.

System name(s)

SalesForce / Azure

Custom-developed code

Yes

Public source code

Not reported

Sourcing

How it was built

Purchased from vendorSource: a) Purchased from a vendor

Vendor (as reported)

Microsoft / SalesForce

Vendors (standardized)

MicrosoftSalesforce

A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.

Data and privacy

Involves PII

Yes

Privacy Impact Assessment

Not reported

Authorization to Operate

Yes

Demographic features

k) None of the above

Training and evaluation data

The processing flow for the Email assistant allows Generative AI to assist our agents in processing email responses to clients from our FHA Resource Center (RC) / contact center. The first step is the FHA RC Rep inputs the email content from the client into the web interface. The system sends the content to the Azure functions on the backend. The backend process the email using Azure Open AI with fined tuned prompts. After the system extracts key information, it classifies use case based on rules and then recommend actions. The results are displayed to the Rep in a organized actional format to review and process the request in a more efficient. The QC process includes review and fine tuning of the technology.

Federal Data Catalog

Not reported

Governance

0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.

Potential impacts and how they were identified

Not reported