AI-Powered Email Extraction and Case Matching Tool

Overview

This project was designed to solve a common operational challenge: inbound emails and copied text often contain customer details buried inside unstructured content. Staff previously had to manually read the text, identify the sender, search for related contacts, and then check whether any existing cases were already open.

I designed an AI-powered Lightning Web Component that extracts email addresses from free text, automatically searches Salesforce contacts, retrieves related cases, and provides an AI-generated summary to help staff triage issues faster.

The challenge

Support teams and service staff frequently receive copied emails, long text threads, and complaint descriptions pasted from Outlook or web forms.

The challenge was that critical information such as:

• Customer email address
• Contact details
• Existing case history
• Issue summary
• Potential duplicate cases

had to be manually identified.

Solution design

Step 1 — intelligent email extraction

The LWC scans pasted text and extracts email addresses using regex logic.

Step 2 — Salesforce contact matching

Once extracted, the component queries Contacts and related Cases using Apex.

Step 3 — AI-assisted summarisation

The raw email body is then passed to an AI summarisation layer that produces a concise issue summary for service officers.

public static List<String> extractEmails(String inputText) {
    Pattern pattern = Pattern.compile(
        '[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Za-z]{2,}'
    );

    Matcher matcher = pattern.matcher(inputText);
    List<String> emails = new List<String>();

    while (matcher.find()) {
        emails.add(matcher.group());
    }

    return emails;
}

This extracted list is then used to search matching Contacts and related Cases.

Implementation example

@AuraEnabled(cacheable=true)
public static List<Contact> searchContactsByEmail(String email) {
    return [
        SELECT Id, Name, Email,
            (SELECT Id, CaseNumber, Status FROM Cases ORDER BY CreatedDate DESC)
        FROM Contact
        WHERE Email = :email
        LIMIT 10
    ];
}

The result is displayed in a structured UI where staff can quickly open the matching contact or related case records.

This dramatically reduced the need to manually navigate across multiple objects.

Approach comparison

Project highlights

Business impact

• Reduced manual triage time
• Improved duplicate case detection
• Faster customer response
• Improved case visibility
• Introduced practical AI into service workflows

FAQ