---
title: "The email comes in. The quote goes out — on its own."
description: "Case study: AI reads the policies and prepares a summary draft for the rep to review. Includes flow screenshots with identifying details removed, and before/after."
canonical: "https://achiya-automation.com/en/case-studies/insurance-policy-extraction/"
last-updated: "2026-10-09"
language: "he"
source: "https://achiya-automation.com"
---

# The email comes in. The quote goes out — on its own.

> Case study: AI reads the policies and prepares a summary draft for the rep to review. Includes flow screenshots with identifying details removed, and before/after.

מקור: https://achiya-automation.com/en/case-studies/insurance-policy-extraction/

---

- [Home](https://achiya-automation.com/en/)
- [Case Studies](https://achiya-automation.com/en/case-studies/)
- From policies in the inbox to a summarized quote on WhatsApp

n8n · Claude AIDocument extractionBased on a project I built

Case study by [Achiya Cohen](https://achiya-automation.com/en/about/), founder of Achiya Automation

At an insurance agency, an email with PDF policies goes through AI extraction for coverages, deductibles, driver age and roadside services. The flow prepares a WhatsApp draft according to the handling rep and records the details in a sheet for review and tracking.
[How it's built](#flows)[See the results](#results)

PDF

Policy-document input

AI

Fields extracted into a draft

Review

Before customer-facing use

Before & After

## Every email requires manual decoding Email comes in → the customer is updated
Before
- Emails with PDF policies come in throughout the day
- A rep opens, reads and extracts by hand: mandatory, comprehensive, third party, deductible
- Drafting a summary for the customer adds work to every quote
- Peak-hour load = customers waiting hours for an answer

After
- AI reads the policies in the email — including tables
- Extracts coverages, deductibles, driver age and seniority into a reviewable draft
- Builds a summary draft according to the handling rep
- The quote details and status are recorded in a tracking sheet

The flow itself

## Here's how it works inside

A description of each flow, with identifying details removed.
FLOW · 01

### Policy email → AI extraction → WhatsApp + logging

The email is captured and the attached policies are read by an AI model that extracts predefined fields: mandatory vs. comprehensive, third party, deductible, the young driver's age and seniority. The system identifies the handling rep by email address, prepares a Hebrew summary draft and records the quote details in the agency's tracking sheet.
Email triggerPDF extractionClaude AIWhatsApp templateGoogle Sheets

**→2**👀→**3**🤖→**4**📊→**5**📲

Results

## The numbers you feel every day
Structured

Field extraction

The required fields are defined in the flow

Draft

Customer message

The wording is prepared for the handling rep to review

Tracking

Quote records

Details and status are kept in the sheet

## Are documents decoded by hand at your company too?

Policies, contracts, quotes — if someone reads and copies by hand, there's an automation here waiting to be built.
[Let's talk on WhatsApp](https://wa.me/972505377060?text=Hi%20Achiya%2C%20I%20came%20from%20the%20website)[Leave your details](https://achiya-automation.com/en/#contact)

Want a system like this? Check out our [WhatsApp bot service](https://achiya-automation.com/en/services/whatsapp-bot/) or our [pricing & plans](https://achiya-automation.com/en/pricing/).

---

סיכום האתר לסוכנים: https://achiya-automation.com/llms.txt · כל העמודים: https://achiya-automation.com/sitemap-index.xml · API: https://achiya-automation.com/openapi.json