Precise PII Detection at Scale

Self-hosted, air-gapped PII detection with enterprise-grade security.

60+

Entity Types

15+

Supported Countries

1M+

Token Context

Core Operations

PII Eraser offers four modes of operation to detect and anonymize PII, PCI and other sensitive data.

Detect (NER)

Identify selected entities, returning character offsets and confidence scores. Supports overlapping and nested PII.

Redact

Replace entities with semantic placeholders like <NAME>. Ideal for sanitizing LLM chats or training data.

Mask

Obfuscate characters while preserving format length (e.g., John → ####). Essential for PCI DSS and ASR transcript readability.

Hash

Replace entities with SHA-256 or SHA-512 hashes. Enables unique user analytics without exposing raw identifiers.

Python Input

import requests

# Detect sensitive data instances
response = requests.post(
    "http://localhost:8000/text/detect",
    json={
        "text": ["Hallo Matthias"],
    }
)

import requests

# Replace sensitive data with entity types
response = requests.post(
    "http://localhost:8000/text/transform",
    json={
        "text": ["I live at Berlinerstr 34, Magdeburg"],
        "operator": "redact", 
    }
)

import requests

# Mask characters with asterisks
response = requests.post(
    "http://localhost:8000/text/transform",
    json={
        "text": ["My Nino is QQ689643C"],
        "operator": "mask", 
    }
)

import requests

# Cryptographic hashing for analytics
response = requests.post(
    "http://localhost:8000/text/transform",
    json={
        "text": ["Medicare no. 3278851195"],
        "operator": "hash", 
    }
)

Output JSON

response.json()

{
  "entities": [
    [
      {
        "entity_type": "NAME",
        "start": 6,
        "end": 14,
        "score": 0.995
      }
    ]
  ],
  "stats": { "total_tokens": 4, "tps": 5420 }
}

{
  "text": [
      "I live at <ADDRESS>"
  ],
  "entities": [
      [
          {
              "entity_type": "ADDRESS",
              "output_start": 10,
              "output_end": 19
          }
      ]
  ],
  "stats": {"total_tokens": 13, "tps": 4550.87}
}

{
  "text": [
      "My Nino is #########"
  ],
  "entities": [
      [
          {
              "entity_type": "GOV_SERVICES_ID",
              "output_start": 11,
              "output_end": 20
          }
      ]
  ],
  "stats": {"total_tokens": 11, "tps": 4282.22}
}

{
  "text": [
      "Medicare no. 041827ef234aeddd34e45a4eb9ef23240fdd7072824a372bb0953377557c6d9c"
  ],
  "entities": [
      [
          {
              "entity_type": "HEALTHCARE_ID",
              "output_start": 13,
              "output_end": 77
          }
      ]
  ],
  "stats": {"total_tokens": 10, "tps": 4640.52}
}

For the full range of options, including how to configure entity types via YAML, please visit the documentation

Designed for the Agentic AI Era

PII Eraser natively understands OpenAI-format chat conversations, delivering context-aware detection for LLM guardrail and agentic workflows.

Existing PII Guardrails

Fast

Low

Scans messages individually. Fast, but misses PII requiring context.

Hi, I need to update the beneficiary details for the 'Project Alpha' contract payouts.

I can help with that. Which specific banking detail do you need to amend?

The bank account number has changed for our UK entity.

Understood. Please provide the new 8-digit account number.

82910453

Process All Messages

Slow

High

Scans entire history every turn. High accuracy, but scales poorly.

Hi, I need to update the beneficiary details for the 'Project Alpha' contract payouts.

I can help with that. Which specific banking detail do you need to amend?

The bank account number has changed for our UK entity.

Understood. Please provide the new 8-digit account number.

<BANK_ACCT>

Smart Context

Fast

High

Automatically includes relevant context. The optimal balance.

Hi, I need to update the beneficiary details for the 'Project Alpha' contract payouts.

I can help with that. Which specific banking detail do you need to amend?

The bank account number has changed for our UK entity.

Understood. Please provide the new 8-digit account number.

<BANK_ACCT>

OpenAI Chat Support

Natively process OpenAI-format chats with context pooled between messages for improved accuracy.

Selective Role Processing

Choose to anonymize only user messages while keeping system prompts and assistant responses intact.

Incremental Processing

Optimize latency-sensitive apps by skipping previously processed messages while maintaining full conversational context.

Great Accuracy, Globally

Accurate identification of 60+ entity types across Western Europe, North America and Australia.

60+ Localized Entity Types

Most tools only detect universal identifiers like emails & phone numbers. PII Eraser features deep, country-level localizations, with support for German Steuer-IDs, French NIR numbers, Australian TFNs and dozens more.

Regular Model Updates

The world changes fast. Older models fail on terms like "COVID" and aren't familiar with MCP tool calls. We continuously update our models to recognize contemporary entities and the shifting GenAI landscape.

No Regex Maintenance

PII Eraser relies on large encoder transformer models, freeing your team from maintaining fragile regex-based solutions. We also offer model updates free of charge in case we do miss something.

Protect Confidential Business Information

Detect and remove customer names, deal values, project codenames, and internal identifiers that create risk even when they aren't covered by privacy regulations.

60+ Entities

System Architecture

Customer Infrastructure (VPC / On-Prem)

Self-hosted, air-gapped deployment. Your data never leaves your environment.

Internal Network /
API Gateway

Text, Chats, Documents

REST API

Text or Chat

PII Eraser Containers

Hardened Container

Chainguard & minimal dependencies

Fast CPU Inference

AVX-512 VNNI & AMX optimizations

Horizontal Scaling

Scale on CPU behind a load balancer

JSON RESPONSE

De-identified Text or Chat

Downstream
Systems

LLMs, RAG, Analytics

Optional: Config File Allow lists, operators & entity types

High Throughput

Sustain >5,000 tokens/sec on standard 8 vCPU instances via AVX-512 VNNI and AMX instruction sets.

Bank-Grade Deployment

Multiple reference implementations including AWS Fargate & ECS and Kubernetes. Deploy a hardened, distroless container built for zero-CVE compliance.

Seamless Migration

PII Eraser provides drop-in compatibility for Microsoft Presidio Analyzer.

Why PII Eraser?

See how a purpose-built, self-hosted solution compares to the alternatives.

	PII Eraser	Cloud API	Open Source	LLM (Generative)
Data Sovereignty	100% Local / Air-gapped	Cloud Only	Local	Cloud (Mostly)
Cost Model	Fixed Price Unlimited	Per Character (Expensive)	Free (Maintenance Heavy)	Per Token (Very Expensive)
EU Localization	Native	Limited	Limited	Native
Native Chat Support	OpenAI Completions Format	No	No	No Structured Chat Input
Long Input Support	1M+ Tokens, No Accuracy Loss	Supported	Chunking Required; Accuracy Degrades	Accuracy Degrades with Length
Latency	Real-time	Medium	Medium	High (> 1000ms)
Security & Supply Chain	Hardened, Minimal Dependencies	Provider Managed	Self-Managed Security & Patching	Provider Managed
Hallucinations	Zero	Zero	Zero	Possible (Probabilistic)

See detailed comparison →

Precise PII Detection at Scale

Core Operations

Designed for the Agentic AI Era

Existing PII Guardrails

Process All Messages

Smart Context

Great Accuracy, Globally

System Architecture

Customer Infrastructure (VPC / On-Prem)

High Throughput

Bank-Grade Deployment

Seamless Migration

Why PII Eraser?

Ready to see it in action?