For Technical Teams

Built for Developers, by Developers

Production AI governance. Open-source SDKs. Built and run by the team that ships it.

Tork
GOVERNANCE

AI Agent Governance, Built for Production

Enterprise-grade oversight for autonomous AI systems

PII Redaction & Data Protection

Automatic detection and masking of sensitive data before it reaches AI models

Policy Enforcement Engine

Define and enforce organizational AI usage policies at the gateway level

Compliance Receipts & Audit Trails

Immutable logs of all AI interactions for regulatory compliance

TORKING-X Metrics Standard

Standardized measurement framework for AI agent performance and safety

Human-in-the-Loop (HITL) Controls

Configurable approval workflows for high-risk AI operations

MCP Gateway Integration

Native Model Context Protocol support for seamless tool orchestration

Circuit Breakers & Kill Switch

Automatic failsafes and manual override capabilities for AI agents

example.py
# pip install tork-governance
from tork_governance import Tork

tork = Tork()

# Detect and redact PII before it reaches the model
result = tork.govern("My TFN is 123 456 789 and email is jane@example.com")

print(result.output)             # PII replaced with redaction tokens
print(result.pii.types)          # ['tfn', 'email']
print(result.receipt.receipt_id) # Cryptographic compliance receipt

# Region- and industry-specific detection
result = tork.govern(
    "Medicare: 2123 45670 1",
    region=["au"],
    industry="healthcare"
)

SDKs & Integration

Python SDK

tork-governance

$ pip install tork-governance
View on PyPI

MCP Server

@torknetwork/mcp-server

$ npm install @torknetwork/mcp-server
View on npm

Framework Adapters

Framework adapters for LangChain, CrewAI, AutoGen, Semantic Kernel and more.

LangChainCrewAIAutoGenSemantic KernelOpenAI AgentsAnthropic Claude

Tech Stack

Next.js 15React 19TypeScriptPython 3.11+FastAPISupabasePostgreSQLVercel EdgeOpenAI APIAnthropic API

Enterprise Features

SSO & RBAC

Single Sign-On with Role-Based Access Control

Multi-tenant Architecture

Isolated environments for each organization

Dedicated Support

Priority technical support channel

Custom Deployment

On-premise or private cloud options

Ready to Integrate?

Talk to our technical team about your architecture requirements and integration needs.

Talk to Our Technical Team