iscover 
Finspo’s new tool extracts client data from identification documents, pre-populates application fields, and checks the information against existing records, a shortcut that the brokerage claims cuts around 20 minutes of manual work from each application.
The feature is the latest release under Finspo’s Automated Data Ingestion (ADI) program, which aims to reduce manual data entry by pulling client information directly from documents and third-party data sources.
How it works
Clients upload identification documents including Australian birth certificates, driver’s licences, passports, and selected overseas passports through Finspo’s online portal.
The AI then reads the documents and extracts the relevant information before checking it against data already held on file.
Where the details match, the field is verified without a manual cross-check. If there is a discrepancy, the system flags it for a broker or administration team member to review.
The extracted data and identification documents remain visible to Finspo staff in the context of the overall application, allowing for human oversight throughout the submission process.
The brokerage built the system in-house, combining AI models for image and text processing with hundreds of internal business rules to check extracted information before it is trusted.
The tool was developed around the identification documents most commonly used by Finspo clients, including Australian ID formats and selected overseas passports.
Finspo said the system is achieving 99 per cent accuracy across the document types its clients most commonly use, with the team refining the underlying models and rules when errors are identified.
Not a replacement for identity checks
Finspo said that the system was not intended as a replacement for lenders’ Know Your Customer (KYC) identity checks.
Instead, it is designed to ensure the information entered into a loan application matches the supporting identification documents before the application is submitted to a lender.
Finspo CEO and co-founder Angus Gilfillan said the brokerage chose to build the system in-house because of the importance of data verification in the home loan process.
“ID data verification is such a critical part of the home loan process that we wanted to be certain of it, and in control of exactly how the AI works at this step, so we built our own, using leading AI models and hundreds of rules our product team has refined over time,” he said.
“It’s already saving our team 20 minutes per application, and it’s making the verification our brokers do every day more accurate, not less.”
Data security
Finspo said that the tool ran on its own infrastructure and that real client data was not used to train the models underpinning the system.
It also added that document data was stored and encrypted in line with industry-standard security practices.
Further development underway
The rollout is the third release under Finspo’s ADI roadmap, following an initial proof-of-concept phase and an earlier upgrade to how ID documents are captured and fed into its mortgage broking platform.
The next stage will extend the automation to bank statements and asset and liability checks through open banking.
AI-powered payslip reading and automated expense categorisation are also being considered further down the track.
[Related: APRA and ASIC roundtables urge action on frontier AI]
Want to see more stories from trusted news sources?Make Broker Daily a preferred news source on Google.
