Use Case: PII Identification & Extraction

Identifying and extracting PII from documents

This video demonstrates how to use Noticia Go’s AI review tool to identify and extract personally identifiable information (PII) from documents. It begins with a focused prompt designed to extract names and email addresses in a structured format, which is helpful for breach notification workflows. The walkthrough also includes a broader prompt to detect multiple types of PII such as Social Security Numbers, dates of birth, and financial account details. These approaches support effective PII detection, tracking, and reporting.

How Noticia Go helps with PII identification

Noticia Go applies AI to accurately detect and extract PII from unstructured documents. Prompts can be tailored for targeted identification (e.g., names and emails) or comprehensive analysis across a broad range of PII categories. Output is structured to support reporting, auditing, and regulatory compliance efforts. This reduces the need for manual data searches and ensures consistent treatment of sensitive information.

Sample PII identification prompts

Prompt 1:

Identify all names and email addresses in this document. Pair them in the format: [Name] <[email]> // [Name] <[email]>. Only include names that are directly associated with email addresses, and only include pairs where two people are mentioned together (e.g., in to/from/cc lines or conversational context). Separate each pair with a double-slash "//".

Prompt 2:

Thoroughly scan this document to identify all instances of Personally Identifiable Information (PII). Consider a broad range of PII types including, but not limited to: full names, dates of birth, Social Security Numbers (or other national identifiers), passport numbers, driver's license numbers, physical addresses, email addresses, phone numbers, financial account numbers (bank accounts, credit cards), medical record numbers, health information, biometric data, IP addresses, geolocation data, and any other information that could be used to identify an individual.

Please provide your findings structured as follows:

PII Presence: [Clearly state PII PRESENT or NO PII DETECTED]
PII Types Found: [If PII is present, provide a comprehensive list of all distinct types of PII identified in the document (e.g., FULL NAME, EMAIL ADDRESS, SOCIAL SECURITY NUMBER, BANK ACCOUNT NUMBER, DATE OF BIRTH). If no PII is detected, state "N/A".]
Detailed Summary and Examples of PII: [If PII is present, provide a detailed summary. For each type of PII found, indicate the approximate count of instances. Include a few specific, representative examples for each category, redacting partially if appropriate for demonstration (e.g., "Social Security Numbers (3 instances, e.g., XXX-XX-1234)", "Email Addresses (5 instances, e.g., john.doe@example.com)", "Bank Account Numbers (1 instance, e.g., ...XXXX1234)"). If occurrences are extremely numerous for a type, note that (e.g., "Full Names (Numerous instances throughout document)"). If no PII is detected, state "N/A".]