Extract data from life insurance applications
Unstract reads life, health and disability income applications, faxed from an agent, scanned or e-signed, and returns checked, structured data to your policy administration system.
waiver rider added per client, Part B exam booked 11/3 — TO
EC-APP-100 (07/2024) · PAGE 1 OF 9
{ "form": "Life insurance application", "application_no": "EC-2611-40877", "proposed_insured": "Marisol K. Trevino", "date_of_birth": "04/17/1984", "ssn": "***-**-6218", "plan": "20-Year Level Term", "face_amount": 750000, "primary_beneficiary": "Daniel R. Trevino", "health_explanation": "ACL repair, right knee, 03/2022", "confidence": 0.95}
Trusted by insurance teams around the globe
A new business inbox full of paper, faxes and e-app printouts
An e-application submitted straight from the agent's laptop reads cleanly. The rest of the new business queue looks like this.
Signed at the client's kitchen table, faxed from the agency
State versions of one application, each with its own question order
A yes to a health question, explained by hand in the margin
Application, HIPAA authorization, replacement notice and illustration together
Small employer census sheets beside the enrollment forms
Your new business team keys each one by hand before an underwriter can order a single requirement, and the applicant waits on a decision the agent promised in days.
Unstract reads every application, pulls the insured, coverage, beneficiary and health answers, checks them against each other and posts the case to your policy administration system.
Read, extract and route every new life case
LLMWhisperer parses the application, Document Studio extracts the fields and Flow runs the new business intake. Each product stands on its own; use one today and connect the others later.
Turn faxed and e-signed life insurance applications into layout-preserved text
A life application is mostly yes/no boxes, a beneficiary table and a few lines of handwriting. LLMWhisperer keeps each tick beside its question and each explanation beside the answer it explains, so a yes on question 7 never drifts into question 8 and a contingent beneficiary's share stays on the contingent line.
Pass the text to your own LLM, or make it the first step of your new business intake.
Learn more about LLMWhispererGet structured data from life and health insurance applications
Name the fields underwriting needs, from date of birth and face amount to each health answer and its explanation, and every application returns the same JSON record whatever carrier form or state version it came on. The same schema covers health insurance application data extraction for individual and small employer plans. A value that contradicts another page in the case arrives flagged, with the page it conflicts with.
Call it as an API from your e-app platform, run it as an ETL pipeline over a scanned mail folder, or make it one step in an orchestrated intake.
Learn more about Document StudioRun new business intake from received application to underwriting decision
Describe your life application intake in plain English and Flow's agents build the workflow: applications are collected from the e-app feed, agent faxes and the mailroom, checked against pending cases and in-force policies, read, and compared with your underwriting rules for age, face amount and health answers. Clean cases go on to accelerated underwriting, and the data lands in your underwriting workbench and your policy administration system, such as Guidewire PolicyCenter or Duck Creek Policy. You approve the workflow once, and it runs the same way on every application.
Set the conditions that need a person, such as a face amount above your automatic limit, a yes on a health question, a missing HIPAA authorization or a low-confidence read, and only those cases stop for an underwriter, with the reason on the case.
Learn more about FlowStart with extraction on a handful of applications
Start for free covers parsing and extraction. Automating your new business intake begins with a consultation.
Life, health and disability applications, and the forms that travel with them
Documents
- Individual life insurance applications
- Term and permanent life applications
- Personal health insurance applications
- Small employer health insurance applications
- Life and disability income insurance applications
- Health insurance change forms
- Beneficiary designation and change forms
- HIPAA and medical information authorizations
- Replacement notices
- Agent reports and producer statements
- Paramedical exam forms
- Attending physician statements
- Employee census and group enrollment forms
- Policy change and reinstatement requests
- Aviation, avocation and tobacco questionnaires
- Other carrier applications and state versions
Fields captured
Processes these applications feed
Built for the people who set up and decide new cases
Applications set up as cases the day they arrive, with missing forms and blank questions listed before anyone chases the agent.
Health answers, handwritten explanations and tobacco use laid out beside the face amount, so each case opens ready for a decision.
Paper and faxed applications checked for missing signatures and unanswered questions before they reach the carrier, not after it sends them back.
Small group enrollment, census and change forms read in one pass, with dependents and plan tiers captured for each employee.
Less keying in new business, quicker decisions for applicants
Handled like the medical file it is
A life application holds a Social Security number, income, the applicant's medical history in their own words and the names of the people who inherit the benefit. The health answers are HIPAA-grade health information, and the rest is enough to open credit in the applicant's name.
- Encryption in transit and at rest for every page of the application.
- Applications you send are never used to train models.
- Access by role and team, so health answers reach underwriters and not the whole office, with each correction logged against the case.
- Retention on your terms: application files are deleted on the schedule you set.
- Self-hosted deployment for insurers that keep applicant data inside their own environment.
What new business and underwriting teams ask first
Reading the applications, authorizations, physician statements and change forms that arrive with a life case, pulling the fields underwriting needs, and setting the case up without anyone retyping it. Unstract does it in three layers: LLMWhisperer parses each page, Document Studio extracts the fields, and Flow runs the intake around them.
Yes. The extracted record carries the proposed insured, coverage, beneficiaries and health answers, so it can prefill your underwriting workbench and your policy administration system instead of a person keying it. Requirements such as an exam or an attending physician statement can be ordered from the same record.
It sits in front of it. Your document management system keeps storing the file; Unstract reads it on the way in, so each stored application arrives with its data already extracted and indexed against the case.
Yes. Life insurance application OCR from LLMWhisperer reads printed and handwritten text together and keeps the layout, so a ticked yes and the explanation written beside it stay attached to the right question. Anything it reads with low confidence goes to a person with the page in view.
Yes. Document Studio deploys an endpoint for your application schema: send the PDF and get the record back as JSON. Teams with an e-app platform call it directly; teams with a scanning room run it as an ETL pipeline over a folder or S3 bucket.
Personal and small employer health insurance applications, life and disability income applications, health insurance change forms, beneficiary changes and reinstatement requests. Fields are found by what they mean rather than where they sit, so state versions and older carrier forms come out in the same record.
Yes. The application, HIPAA authorization, replacement notice, illustration and agent report are separated first, then each is read on its own and the values are compared across them, such as the face amount on the application against the illustration.
The case stops in a review queue with both values and the pages they came from. An underwriter confirms, corrects or orders a requirement, the decision is recorded against the case, and the rest of the queue keeps moving.
Yes. Describe the intake in plain English and Flow's agents build the workflow: collect applications, parse, extract, check against your underwriting rules and post the case. You review and approve it before it runs, and it then runs identically on every case, stopping for an underwriter only where you told it to.
Unstract is SOC 2, ISO 27001, GDPR and HIPAA compliant. Applications are encrypted in transit and at rest, never used to train models and kept only as long as you choose, and insurers can self-host. Details are in the Trust Center.
To test extraction today, start for free and upload a few applications. For the full intake, a consultation maps your forms, your underwriting rules and your policy administration system, and most teams have a first live run within days.
Bring last week's new business applications
Life, health and disability applications, faxed and handwritten ones included. We run them live and show every field, check and referral.