How We Protect Your Data

How our browser-side processing and privacy-first architecture protects your client data in the IFA Letter Generator.

Privacy Architecture

✓Browser-Side Document Processing

When you upload a meeting note (PDF or Word), it is processed directly in your browser. The original document is not uploaded to our servers; only extracted text is sent securely for analysis.

✓Transient Processing

Meeting note text is processed transiently. It is not stored after AI extraction. Only structured fields (template type, fund names, adviser details) are retained in the database.

✓Active: PII Masking Gateway

Before any text reaches the AI model, a browser-side PII masking layer automatically detects and replaces personal identifiers with neutral placeholders. Client names, addresses, NI numbers, postcodes, email addresses, phone numbers, dates of birth, and account references are all masked. After AI extraction, original values are reinjected into the results. So the AI never sees the underlying personal details.

Active: PII Masking Categories

The PII masking gateway is active. The following categories are automatically detected and masked before AI processing:

The gateway uses pattern-based rules tuned for UK financial services data to detect and replace:

Data Type	Will Be Replaced With
Client full names	[CLIENT_NAME]
Street addresses	[ADDRESS_REDACTED]
National Insurance numbers	[NI_REDACTED]
Dates of birth	[DOB_REDACTED]
Account / policy references	[ACCOUNT_REDACTED]
Email addresses	[EMAIL_REDACTED]
Phone numbers	[PHONE_REDACTED]
UK postcodes	[POSTCODE_REDACTED]

Data Handling Best Practices

With PII masking active, the following best practices remain recommended:

Review meeting notes before upload and redact any unnecessary personal information where possible.
The system processes meeting note text transiently. Raw text is not stored after AI extraction.
Only structured fields (template type, fund names, adviser details) are retained in the database.
All data is encrypted at rest (AES-256) and in transit (TLS 1.2+).
Row-level security ensures each firm's data is fully partitioned from other firms.

How Can You Be Sure?

We understand that "trust us, it's local" isn't enough. Especially when dealing with client financial data. Here's a plain-English explanation of why browser-side processing is verifiable.

The photocopier analogy

Think of it like a photocopier with a shredder built in. When you load a letter into a photocopier, the machine reads it and produces a copy. But the original letter never leaves the room. The IFA Letter Generator works the same way. Your meeting note is "read" inside your own browser. The system extracts the words, then the original document stays put. Only the extracted words are ever sent anywhere.

The technical reason

Document processing is performed using standard browser technologies. It runs entirely within the browser tab. No server connection is made during this step. This is not a policy choice; it is how the technology is architected.

What an IT team can verify

No outbound file upload request during the document loading step
The first outbound request only occurs after clicking "Upload"
That request contains plain text. Not a file, and not binary data

In plain terms: Your meeting note document never travels across the internet. The words inside it do. But the document itself stays on your computer at all times.

What This Means For You

Your original documents never leave your browser.
Meeting note text is processed transiently. Not stored after extraction.
All stored data is encrypted at rest and in transit.
Your firm's data is fully partitioned from other firms via row-level security.
PII masking is active. Personal identifiers are automatically replaced with placeholders before AI processing.
You can verify browser-side processing independently using your browser's developer tools.

This document explains our privacy architecture. For the full picture, see:

···

Questions? Contact info@caseflowautomation.co.uk