> ## Documentation Index
> Fetch the complete documentation index at: https://docs.021labs.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# PII Redaction

> Automatically redact sensitive data from traces

## Overview

RDK can automatically redact Personally Identifiable Information (PII) from your traces before they're sent to the collector. This helps you maintain compliance while still getting valuable observability data.

## Quick Start

### Enable built-in redaction

Pass `redact_pii=True` to `init()` to enable default PII redaction:

```python theme={null}
from rdk import init

init(redact_pii=True)
```

### Custom redaction

For fine-grained control, build a redactor with `RedactorConfig`:

```python theme={null}
import re
from rdk import init
from rdk import RedactorConfig, create_redactor

redactor = create_redactor(RedactorConfig(
    redact_emails=True,
    redact_phones=True,
    redact_ssn=True,
    redact_credit_cards=True,
    redact_api_keys=True,
    custom_patterns=[
        (re.compile(r"ACC-\d{8}"), "[ACCOUNT REDACTED]"),
        (re.compile(r"EMP-\d{6}"), "[EMPLOYEE_ID REDACTED]"),
    ],
))

init(redactor=redactor)
```

## Built-in Patterns

| Type        | Example               | Redacted As          |
| ----------- | --------------------- | -------------------- |
| Email       | `john@example.com`    | `[EMAIL REDACTED]`   |
| Phone       | `555-123-4567`        | `[PHONE REDACTED]`   |
| SSN         | `123-45-6789`         | `[SSN REDACTED]`     |
| Credit Card | `4111-1111-1111-1111` | `[CARD REDACTED]`    |
| API Key     | `sk-abc123...`        | `[API_KEY REDACTED]` |

## RedactorConfig Fields

```python theme={null}
class RedactorConfig:
    redact_emails: bool = True
    redact_phones: bool = True
    redact_ssn: bool = True
    redact_credit_cards: bool = True
    redact_api_keys: bool = True
    custom_patterns: list[tuple[re.Pattern, str]] | None = None
    custom_redactor: Callable[[str], str] | None = None
```

<ParamField path="redact_emails" type="boolean" default={true}>
  Redact email addresses.
</ParamField>

<ParamField path="redact_phones" type="boolean" default={true}>
  Redact phone numbers (formats: `555-123-4567`, `555.123.4567`, `5551234567`).
</ParamField>

<ParamField path="redact_ssn" type="boolean" default={true}>
  Redact US Social Security Numbers (`123-45-6789`).
</ParamField>

<ParamField path="redact_credit_cards" type="boolean" default={true}>
  Redact 16-digit credit card numbers.
</ParamField>

<ParamField path="redact_api_keys" type="boolean" default={true}>
  Redact API keys matching common patterns (`sk-`, `api_key`, `bearer ...`).
</ParamField>

<ParamField path="custom_patterns" type="list[tuple[re.Pattern, str]]" default="[]">
  List of `(compiled_pattern, replacement_string)` tuples applied after built-in patterns.

  ```python theme={null}
  import re
  custom_patterns=[
      (re.compile(r"ACC-\d{8}"), "[ACCOUNT REDACTED]"),
  ]
  ```
</ParamField>

<ParamField path="custom_redactor" type="Callable[[str], str]" default="None">
  A function applied to each string value before the built-in patterns. Use for custom logic that regex can't express.
</ParamField>

## Utility Functions

```python theme={null}
from rdk import redact_all_pii
from rdk.filters import create_default_redactor
```

**`redact_all_pii(value)`** — One-shot redaction using all default patterns. Useful for sanitizing data outside of traces.

```python theme={null}
from rdk import redact_all_pii

clean = redact_all_pii({
    "message": "Call me at 555-123-4567 or john@example.com"
})
# {"message": "Call me at [PHONE REDACTED] or [EMAIL REDACTED]"}
```

**`create_default_redactor()`** — Returns a reusable redactor with all defaults enabled.

## What Gets Redacted

Redaction is applied to:

* Span inputs — messages sent to the LLM
* Span outputs — responses from the LLM
* Tool arguments and results

Redaction is **not** applied to:

* Trace/span IDs
* Timestamps
* Token counts
* Model names
* Metadata keys (only values are scanned)

## Best Practices

1. Always enable PII redaction in production
2. Test custom patterns against realistic sample data before deploying
3. Use `custom_redactor` for preprocessing (e.g., tokenization) before regex matching

## See Also

* [create\_redactor()](/api-reference/create-redactor) — API reference
* [RedactorConfig](/api-reference/redactor-config) — Full config reference