piighost

English | Français

piighost replaces personal data with placeholders before the LLM sees it, then restores the real values in the reply and in tool calls. This Python library de-identifies confidential data, meaning personal data (PII) and secrets, through a reversible pseudonymization often called anonymization. LangChain, Pydantic AI, LlamaIndex and Claude Code integrations are provided, together with an OpenAI and Anthropic API connector.

This de-identification spots confidential data with pluggable detectors (regex, NER, LLM) and replaces each value with a placeholder, the token that takes its place. For example:

• John Doe becomes <<PERSON:1>>
• john.doe@example.com becomes <<EMAIL:1>>

This placeholder stays the same from one message to the next with the conversational pipeline, which keeps the mapping between a value and its placeholder across the whole conversation. If john.doe@example.com reappears three messages later, the placeholder is still <<EMAIL:1>>, which lets the LLM follow the thread.

The LLM therefore only receives de-identified text. When it returns placeholders, for example by answering Hello <<PERSON:1>>, piighost replaces them with the real values. The user sees John Doe and never sees the de-identification.

The same mechanism protects agents that call tools. With the LangChain middleware, a tool that needs the real email address receives it in clear, while the LLM that supplies it only writes <<EMAIL:1>>.

The LLM only sees placeholders. The tool receives the real address, the user gets a clear-text reply, and your agent code stays the same.

Quickstart

pip install piighost   # or: uv add piighost

ExactMatchDetector de-identifies a dictionary of known values, with no model to download.

import asyncio

from piighost.components.detector import ExactMatchDetector
from piighost.pipeline import AnonymizationPipeline

detector = ExactMatchDetector({"John Doe": "PERSON", "john.doe@example.com": "EMAIL"})
pipeline = AnonymizationPipeline(detector)

result = asyncio.run(pipeline.anonymize("Write to John Doe at john.doe@example.com."))
print(result.text)  # Write to <<PERSON:1>> at <<EMAIL:1>>.

The Quickstart goes on with a real detector and a conversation.

Go further

• Start: installation, first pipeline, conversational pipeline
• Configure: a pipeline in a TOML file, pattern groups from the catalog, every config key
• Integrate: LangChain, Pydantic AI, LlamaIndex, Claude Code, a remote piighost-api
• Deploy: an API server from a catalog configuration, a production thread pipeline, several instances
• Understand: why de-identify, architecture, security, GDPR compliance, limitations, detection, measured, how it compares
• Upgrade: versions and the move to 2.0

Ecosystem

• piighost.dev: the project website, with the ecosystem and its philosophy
• piighost catalog: reviewed regex groups and ready-made pipeline configurations, pulled by reference
• piighost-api: a server hosting one pipeline behind HTTP, with OpenAI- and Anthropic-compatible proxies
• piighost-chat: an example chat interface with human-in-the-loop
• caviardage: a browser app that redacts a Word deed, with piighost running in Pyodide. Nothing of the document leaves the tab

Project

• Community: Discord to get help, report bugs and request features
• Contributing: contribution guide and bug reports
• License: MIT

If piighost is useful to you, a star on GitHub helps other developers find it.